element_type
stringclasses 4
values | project_name
stringclasses 1
value | uuid
stringlengths 36
36
| name
stringlengths 0
346
| imports
stringlengths 0
2.67k
| structs
stringclasses 761
values | interfaces
stringclasses 22
values | file_location
stringclasses 545
values | code
stringlengths 26
8.07M
| global_vars
stringclasses 7
values | package
stringclasses 124
values | tags
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|
function | openshift/openshift-tests-private | 949a5e4f-7165-4a86-bd5b-39b0b1d68d6f | runSSHCommand | ['"strings"'] | ['User'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func runSSHCommand(server, user string, commands ...string) (string, error) {
// Combine commands into a single string
fullCommand := strings.Join(commands, " ")
sshkey, err := exutil.GetPrivateKey()
o.Expect(err).NotTo(o.HaveOccurred())
sshClient := exutil.SshClient{User: user, Host: server, Port: 22, PrivateKey: sshkey}
return sshClient.RunOutput(fullCommand)
} | apiserverauth | |||
function | openshift/openshift-tests-private | 52bfff18-ac3c-4481-a8f7-0634afaab821 | getProxyURL | ['"net/url"', '"os"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getProxyURL() *url.URL {
// Prefer https_proxy, fallback to http_proxy
proxyURLString := os.Getenv("https_proxy")
if proxyURLString == "" {
proxyURLString = os.Getenv("http_proxy")
}
if proxyURLString == "" {
return nil
}
proxyURL, err := url.Parse(proxyURLString)
if err != nil {
e2e.Failf("error parsing proxy URL: %v", err)
}
return proxyURL
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 546337b4-98f0-4934-af7b-120276b46266 | getMicroshiftHostname | ['"net/url"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getMicroshiftHostname(oc *exutil.CLI) string {
microShiftURL, err := oc.AsAdmin().WithoutNamespace().Run("config").Args("view", "-ojsonpath={.clusters[0].cluster.server}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
fqdnName, err := url.Parse(microShiftURL)
o.Expect(err).NotTo(o.HaveOccurred())
return fqdnName.Hostname()
} | apiserverauth | ||||
function | openshift/openshift-tests-private | c2de446f-6683-4d4d-a9ee-b2908a84879f | applyLabel | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func applyLabel(oc *exutil.CLI, asAdmin bool, withoutNamespace bool, parameters ...string) {
_, err := doAction(oc, "label", asAdmin, withoutNamespace, parameters...)
o.Expect(err).NotTo(o.HaveOccurred(), "Adding label to the namespace failed")
} | apiserverauth | |||||
function | openshift/openshift-tests-private | 9cb9226a-84cb-4be9-aedb-518f784d2609 | checkUserAuditLog | ['"context"', '"fmt"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func checkUserAuditLog(oc *exutil.CLI, logGroup string, user string, pass string) (string, int) {
var (
eventLogs string
eventCount = 0
n int
now = time.Now().UTC().Unix()
)
errUser := oc.AsAdmin().WithoutNamespace().Run("login").Args("-u", user, "-p", pass).NotShowInfo().Execute()
o.Expect(errUser).NotTo(o.HaveOccurred())
whoami, err := oc.AsAdmin().WithoutNamespace().Run("whoami").Args("").Output()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("whoami: %s", whoami)
err = oc.AsAdmin().WithoutKubeconf().WithoutNamespace().Run("logout").Args().Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("The user %s logged out successfully", user)
script := fmt.Sprintf(`rm -if /tmp/audit-test-*.json;
for logpath in kube-apiserver oauth-apiserver openshift-apiserver;do
grep -h "%s" /var/log/${logpath}/audit*.log | jq -c 'select (.requestReceivedTimestamp | .[0:19] + "Z" | fromdateiso8601 > %v)' >> /tmp/audit-test-$logpath.json;
done;
cat /tmp/audit-test-*.json`, logGroup, now)
contextErr := oc.AsAdmin().WithoutNamespace().Run("config").Args("use-context", "admin").Execute()
o.Expect(contextErr).NotTo(o.HaveOccurred())
e2e.Logf("Get all master nodes.")
masterNodes, getAllMasterNodesErr := exutil.GetClusterNodesBy(oc, "master")
o.Expect(getAllMasterNodesErr).NotTo(o.HaveOccurred())
o.Expect(masterNodes).NotTo(o.BeEmpty())
for _, masterNode := range masterNodes {
eventLogs, n = checkAuditLogs(oc, script, masterNode, "openshift-kube-apiserver")
e2e.Logf("event logs count:%v", n)
eventCount += n
}
return eventLogs, eventCount
} | apiserverauth | ||||
function | openshift/openshift-tests-private | a41c3f78-6790-484c-9cd9-d235c132018b | verifyMicroshiftLogs | ['"context"', '"fmt"', '"strings"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func verifyMicroshiftLogs(nodename string, cmd string, cmp string) (string, int, error) {
var (
output string
err error
)
mstatusErr := wait.PollUntilContextTimeout(context.Background(), 6*time.Second, 200*time.Second, false, func(cxt context.Context) (bool, error) {
output, err = runSSHCommand(nodename, "redhat", cmd)
if err != nil {
return false, err
}
count := len(strings.TrimSpace(output))
switch cmp {
case "==":
if count == 0 {
return true, nil
}
case ">":
if count > 0 {
return true, nil
}
case "<":
if count < 0 {
return true, nil
}
default:
return false, fmt.Errorf("invalid comparison operator")
}
return false, nil
})
return output, len(strings.TrimSpace(output)), mstatusErr
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 14cdf7ff-f604-4333-a256-43f5ef9cac51 | getMicroshiftConfig | ['"context"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"', '"github.com/tidwall/gjson"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getMicroshiftConfig(nodeName string, cmd string, keyValue string) (string, error) {
var strValue string
mstatusErr := wait.PollUntilContextTimeout(context.Background(), 30*time.Second, 300*time.Second, false, func(cxt context.Context) (bool, error) {
// Run SSH command to get the YAML configuration
yamlData, err := runSSHCommand(nodeName, "redhat", cmd)
if err == nil && yamlData != "" {
yamlToJson, yamlErr := exutil.Yaml2Json(yamlData)
if yamlErr == nil && yamlToJson != "" {
// Parse YAML data
yamlJson := gjson.Parse(yamlToJson).String()
if yamlJson != "" {
// Get value from JSON using provided key
strValue = gjson.Get(yamlJson, keyValue).String()
if strValue != "" {
e2e.Logf("Config values : %s", strValue)
return true, nil
}
}
}
}
return false, nil
})
return strValue, mstatusErr
} | apiserverauth | ||||
function | openshift/openshift-tests-private | aa37f676-c80c-4bc4-ac8a-1e6dd28d127a | gatherSosreports | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func gatherSosreports(fqdnName string, user string, sosReportCmd string, tmpdir string) string {
sosreportStatus, sosErr := runSSHCommand(fqdnName, user, sosReportCmd)
o.Expect(sosErr).NotTo(o.HaveOccurred())
e2e.Logf("SOS Report :: %v", sosreportStatus)
o.Expect(strings.Contains(sosreportStatus, "Your sos report has been generated and saved in")).To(o.BeTrue())
o.Expect(strings.Contains(sosreportStatus, tmpdir+"/sosreport")).To(o.BeTrue())
return sosreportStatus
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 4c10530c-05a0-4841-85fa-bb106984238c | clusterSanityCheck | ['"crypto/tls"', '"crypto/x509"', '"fmt"', '"net"', '"net/http"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func clusterSanityCheck(oc *exutil.CLI) error {
var (
project_ns = exutil.GetRandomString()
errCreateProj error
)
statusNode, errNode := getResource(oc, asAdmin, withoutNamespace, "node")
if errNode != nil {
e2e.Logf("Error fetching Node Status: %s :: %s", statusNode, errNode.Error())
if strings.ContainsAny(errNode.Error(), "Unable to connect to the server: net/http: TLS handshake timeout") {
e2e.Failf("Cluster Not accessible, may be env issue issue or network disruption")
}
}
statusCO, errCO := getResource(oc, asAdmin, withoutNamespace, "co")
if errCO != nil {
e2e.Logf("Error fetching Cluster Operators Status: %s :: %s", statusCO, errCO.Error())
if strings.ContainsAny(errCO.Error(), "Unable to connect to the server: tls: failed to verify certificate: x509: certificate signed by unknown authority") {
status, _ := getResource(oc, asAdmin, withoutNamespace, "co", "--insecure-skip-tls-verify")
e2e.Logf("cluster Operators Status :: %s", status)
statusKAS, _ := getResource(oc, asAdmin, withoutNamespace, "co", "kube-apiserver", "-o", "yaml", "--insecure-skip-tls-verify")
e2e.Logf("KAS Operators Status :: %s", statusKAS)
}
}
// retry to create new project to avoid transient ServiceUnavailable of openshift-apiserver
o.Eventually(func() bool {
errCreateProj = oc.AsAdmin().WithoutNamespace().Run("new-project").Args(project_ns, "--skip-config-write").Execute()
return errCreateProj == nil
}, 9*time.Second, 3*time.Second).Should(o.BeTrue(), fmt.Sprintf("Failed to create project %s with error %v", project_ns, errCreateProj))
if errCreateProj != nil && strings.ContainsAny(errCreateProj.Error(), "the server is currently unable to handle the request") {
status, _ := getResource(oc, asAdmin, withoutNamespace, "co")
e2e.Logf("cluster Operators Status :: %s", status)
}
errDeleteProj := oc.AsAdmin().WithoutNamespace().Run("delete").Args("project", project_ns, "--ignore-not-found").Execute()
if errDeleteProj != nil {
e2e.Logf("Error deleting project %s: %s", project_ns, errDeleteProj.Error())
}
if errCO != nil || errCreateProj != nil || errDeleteProj != nil {
return fmt.Errorf("cluster sanity check failed")
}
e2e.Logf("Cluster sanity check passed")
return nil
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 5e4344d0-7fe1-4166-a386-befbee7789f1 | clusterSanityCheckMicroShift | ['"fmt"', '"net"', '"net/http"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func clusterSanityCheckMicroShift(oc *exutil.CLI) error {
statusNode, errNode := getResource(oc, asAdmin, withoutNamespace, "node")
if errNode != nil {
e2e.Logf("Error fetching Node Status: %s :: %s", statusNode, errNode.Error())
if strings.ContainsAny(errNode.Error(), "Unable to connect to the server: net/http: TLS handshake timeout") {
e2e.Failf("Cluster Not accessible, may be env issue issue or network disruption")
}
}
project_ns := exutil.GetRandomString()
errCreateNs := oc.AsAdmin().WithoutNamespace().Run("create").Args("ns", project_ns).Execute()
if errCreateNs != nil {
e2e.Logf("Error creating project %s: %s", project_ns, errCreateNs.Error())
}
errDeleteNs := oc.WithoutNamespace().Run("delete").Args("ns", project_ns, "--ignore-not-found").Execute()
if errDeleteNs != nil {
e2e.Logf("Error deleting project %s: %s", project_ns, errDeleteNs.Error())
}
if errCreateNs != nil || errDeleteNs != nil {
return fmt.Errorf("Cluster sanity check failed")
}
e2e.Logf("Cluster sanity check passed")
return nil
} | apiserverauth | ||||
function | openshift/openshift-tests-private | d6ff4bb0-9234-4caa-82fc-39c238cb46eb | getPendingCSRs | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getPendingCSRs(oc *exutil.CLI) ([]string, error) {
output := getResourceToBeReady(oc, asAdmin, withoutNamespace, "csr")
o.Expect(output).NotTo(o.BeEmpty())
// Convert the output to a string and split it into lines
outputStr := string(output)
lines := strings.Split(outputStr, "\n")
var pendingCSRs []string
// Filter for CSRs with status "Pending" and extract the CSR name
for _, line := range lines {
if strings.Contains(line, "Pending") {
fields := strings.Fields(line)
if len(fields) > 0 {
pendingCSRs = append(pendingCSRs, fields[0]) // Append CSR name to the list
}
}
}
// If no pending CSRs were found, return an empty list and no error
return pendingCSRs, nil
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 7e55f04f-e5c2-4e8c-a437-9775e1d7e096 | getResourceWithKubeconfig | ['"context"', '"fmt"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getResourceWithKubeconfig(oc *exutil.CLI, newKubeconfig string, waitForError bool, getResource ...string) (string, error) {
var output string
var err error
args := append([]string{newKubeconfig}, getResource...)
pollErr := wait.PollUntilContextTimeout(context.Background(), 5*time.Second, 120*time.Second, false, func(ctx context.Context) (bool, error) {
output, err = oc.AsAdmin().WithoutNamespace().WithoutKubeconf().Run("--kubeconfig").Args(args...).Output()
if err != nil {
if waitForError {
return false, nil
}
return true, err
}
return true, nil // Success
})
if pollErr != nil {
if waitForError {
return "", fmt.Errorf("timed out waiting for `%v` command to succeed: %w :: and error is `%v`", getResource, pollErr, err)
}
return "", pollErr
}
return output, err
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 63fc53ff-6933-4770-a9dc-fc5933b5ab81 | kasOperatorCheckForStep | ['"reflect"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func kasOperatorCheckForStep(oc *exutil.CLI, preConfigKasStatus map[string]string, step string, msg string) {
var (
coName = "kube-apiserver"
kubeApiserverCoStatus = map[string]string{"Available": "True", "Progressing": "False", "Degraded": "False"}
)
e2e.Logf("Pre-configuration with %s operator status before %s: %s", coName, msg, preConfigKasStatus)
// It takes about 30 seconds for KAS rolling out from deployment to progress
// wait some bit more time and double check, to ensure it is stably healthy
time.Sleep(45 * time.Second)
postConfigKasStatus := getCoStatus(oc, coName, kubeApiserverCoStatus)
e2e.Logf("Post-configuration with %s operator status after %s %s", coName, msg, postConfigKasStatus)
// Check if KAS operator status is changed after ValidatingWebhook configration creation
if !reflect.DeepEqual(preConfigKasStatus, postConfigKasStatus) {
if reflect.DeepEqual(preConfigKasStatus, kubeApiserverCoStatus) {
// preConfigKasStatus has the same status of kubeApiserverCoStatus, means KAS operator is changed from stable to unstable
e2e.Failf("Test step-%s failed: %s operator are abnormal after %s!", step, coName, msg)
}
}
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 40a3464a-5a54-4f90-822c-25cc5326a4e8 | createSecretsWithQuotaValidation | ['"fmt"', '"strconv"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func createSecretsWithQuotaValidation(oc *exutil.CLI, namespace, clusterQuotaName string, crqLimits map[string]string, caseID string) {
// Step 1: Retrieve current secret count
secretCount, err := oc.Run("get").Args("-n", namespace, "clusterresourcequota", clusterQuotaName, "-o", `jsonpath={.status.namespaces[*].status.used.secrets}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
usedCount, _ := strconv.Atoi(secretCount)
limits, _ := strconv.Atoi(crqLimits["secrets"])
steps := 1
// Step 2: Create secrets and check if quota limit is reached
for i := usedCount; i <= limits; i++ {
secretName := fmt.Sprintf("%v-secret-%d", caseID, steps)
e2e.Logf("Creating secret %s", secretName)
// Attempt to create the secret
output, err := oc.Run("create").Args("-n", namespace, "secret", "generic", secretName).Output()
// Step 3: Expect failure when reaching the quota limit
if i < limits {
output1, _ := oc.Run("get").Args("-n", namespace, "secret").Output()
e2e.Logf("Get total secrets created to debug :: %s", output1)
o.Expect(err).NotTo(o.HaveOccurred()) // Expect success before quota is reached
} else {
// Expect the specific "exceeded quota" error message
if err != nil && strings.Contains(output, "secrets.*forbidden: exceeded quota") {
e2e.Logf("Quota limit reached, as expected.")
} else {
o.Expect(err).To(o.HaveOccurred()) // Fail if any other error occurs
}
}
steps++
}
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 6d66c7a3-a234-44ab-b8c6-a8daf63179c6 | checkDisconnect | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func checkDisconnect(oc *exutil.CLI) bool {
workNode, err := exutil.GetFirstWorkerNode(oc)
o.Expect(err).ShouldNot(o.HaveOccurred())
curlCMD := "curl -I ifconfig.me --connect-timeout 5"
output, err := exutil.DebugNode(oc, workNode, "bash", "-c", curlCMD)
if !strings.Contains(output, "HTTP") || err != nil {
e2e.Logf("Unable to access the public Internet from the cluster.")
return true
}
e2e.Logf("Successfully connected to the public Internet from the cluster.")
return false
} | apiserverauth | ||||
function | openshift/openshift-tests-private | 0a6f1363-f2be-4afc-a5e7-b8e64705316f | validateMicroshiftConfig | ['"regexp"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func validateMicroshiftConfig(fqdnName string, user string, patternToMatch string) {
e2e.Logf("Check manifest config")
chkConfigCmd := `sudo /usr/bin/microshift show-config --mode effective 2>/dev/null`
re := regexp.MustCompile(patternToMatch)
mchkConfig, mchkConfigErr := runSSHCommand(fqdnName, user, chkConfigCmd)
o.Expect(mchkConfigErr).NotTo(o.HaveOccurred())
match := re.MatchString(mchkConfig)
if !match {
e2e.Failf("Config not matched :: \n" + mchkConfig)
}
} | apiserverauth | ||||
function | openshift/openshift-tests-private | eacb3dce-3d74-4a47-a158-415df00509fb | fetchOpenShiftAPIServerCert | ['"context"', '"crypto/tls"', '"encoding/pem"', '"fmt"', '"net"', '"net/http"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func fetchOpenShiftAPIServerCert(apiServerEndpoint string) ([]byte, error) {
timeout := 120 * time.Second
retryInterval := 20 * time.Second
// Create a cancellable context for polling
ctx, cancel := context.WithTimeout(context.Background(), timeout)
defer cancel()
transport := &http.Transport{
TLSClientConfig: &tls.Config{
InsecureSkipVerify: true,
},
}
proxyURL := getProxyURL()
transport.Proxy = http.ProxyURL(proxyURL)
// Set up TLS configuration and DialContext
transport.DialContext = func(ctx context.Context, network, addr string) (net.Conn, error) {
return (&net.Dialer{}).DialContext(ctx, network, addr)
}
client := &http.Client{
Transport: transport,
}
var pemCert []byte
pollFunc := func(ctx context.Context) (done bool, err error) {
// Attempt to send a GET request to the OpenShift API server
resp, err := client.Get(apiServerEndpoint)
if err != nil {
e2e.Logf("Error connecting to the OpenShift API server: %v. Retrying...\n", err)
return false, nil
}
defer resp.Body.Close()
// Check TLS connection state
tlsConnectionState := resp.TLS
if tlsConnectionState == nil {
return false, fmt.Errorf("No TLS connection established")
}
// Encode the server's certificate to PEM format
cert := tlsConnectionState.PeerCertificates[0]
pemCert = pem.EncodeToMemory(&pem.Block{Type: "CERTIFICATE", Bytes: cert.Raw})
if pemCert == nil {
return false, fmt.Errorf("Error encoding certificate to PEM")
}
fmt.Println("Certificate fetched successfully")
return true, nil
}
err := wait.PollUntilContextTimeout(ctx, retryInterval, timeout, true, pollFunc)
if err != nil {
return nil, fmt.Errorf("failed to fetch certificate within timeout: %w", err)
}
return pemCert, nil
} | apiserverauth | ||||
function | openshift/openshift-tests-private | aa0212e8-d50e-4ce3-939b-49e7ae64ae0a | getRandomString | ['"math/rand"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getRandomString(digit int) string {
chars := "abcdefghijklmnopqrstuvwxyz0123456789"
seed := rand.New(rand.NewSource(time.Now().UnixNano()))
buffer := make([]byte, digit)
for index := range buffer {
buffer[index] = chars[seed.Intn(len(chars))]
}
return string(buffer)
} | apiserverauth | ||||
function | openshift/openshift-tests-private | de4e8ccf-d0f5-4f8d-bbf3-d4eca43705d0 | getSAToken | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/apiserverauth/apiserver_util.go | func getSAToken(oc *exutil.CLI, sa, ns string) (string, error) {
e2e.Logf("Getting a token assgined to specific serviceaccount from %s namespace...", ns)
token, err := oc.AsAdmin().WithoutNamespace().Run("create").Args("token", sa, "-n", ns).Output()
if err != nil {
if strings.Contains(token, "unknown command") { // oc client is old version, create token is not supported
e2e.Logf("oc create token is not supported by current client, use oc sa get-token instead")
token, err = oc.AsAdmin().WithoutNamespace().Run("sa").Args("get-token", sa, "-n", ns).Output()
} else {
return "", err
}
}
return token, err
} | apiserverauth | ||||
file | openshift/openshift-tests-private | f871c3b6-41fb-4382-b7cb-01ca4a04e803 | cloudcredential_util | import (
"encoding/base64"
"encoding/json"
"fmt"
"sort"
"strings"
"time"
"k8s.io/apimachinery/pkg/util/wait"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
e2e "k8s.io/kubernetes/test/e2e/framework"
) | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | package cloudcredential
import (
"encoding/base64"
"encoding/json"
"fmt"
"sort"
"strings"
"time"
"k8s.io/apimachinery/pkg/util/wait"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
e2e "k8s.io/kubernetes/test/e2e/framework"
)
const (
ccoNs = "openshift-cloud-credential-operator"
ccoCap = "CloudCredential"
ccoRepo = "cloud-credential-operator"
ccoManifestPath = "manifests"
defaultSTSCloudTokenPath = "/var/run/secrets/kubernetes.io/serviceaccount/token"
DefaultTimeout = 120
)
type prometheusQueryResult struct {
Data struct {
Result []struct {
Metric struct {
Name string `json:"__name__"`
Container string `json:"container"`
Endpoint string `json:"endpoint"`
Instance string `json:"instance"`
Job string `json:"job"`
Mode string `json:"mode"`
Namespace string `json:"namespace"`
Pod string `json:"pod"`
Service string `json:"service"`
} `json:"metric"`
Value []interface{} `json:"value"`
} `json:"result"`
ResultType string `json:"resultType"`
} `json:"data"`
Status string `json:"status"`
}
type credentialsRequest struct {
name string
namespace string
provider string
template string
}
type azureCredential struct {
key string
value string
}
type gcpCredential struct {
key string
value string
}
type OcpClientVerb = string
func doOcpReq(oc *exutil.CLI, verb OcpClientVerb, notEmpty bool, args ...string) string {
e2e.Logf("running command : oc %s %s", string(verb), strings.Join(args, " "))
res, err := oc.AsAdmin().WithoutNamespace().Run(string(verb)).Args(args...).Output()
o.Expect(err).ShouldNot(o.HaveOccurred())
if notEmpty {
o.Expect(res).ShouldNot(o.BeEmpty())
}
return res
}
func getCloudCredentialMode(oc *exutil.CLI) (string, error) {
var (
mode string
iaasPlatform string
rootSecretName string
err error
)
iaasPlatform, err = getIaasPlatform(oc)
if err != nil {
return "", err
}
if iaasPlatform == "none" || iaasPlatform == "baremetal" {
mode = "none" //mode none is for baremetal
return mode, nil
}
//Check if the cloud providers which support Manual mode only
if iaasPlatform == "ibmcloud" || iaasPlatform == "alibabacloud" || iaasPlatform == "nutanix" {
mode = "manual"
return mode, nil
}
modeInCloudCredential, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("cloudcredential", "cluster", "-o=jsonpath={.spec.credentialsMode}").Output()
if err != nil {
return "", err
}
if modeInCloudCredential != "Manual" {
rootSecretName, err = getRootSecretName(oc)
if err != nil {
return "", err
}
modeInSecretAnnotation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", rootSecretName, "-n=kube-system", "-o=jsonpath={.metadata.annotations.cloudcredential\\.openshift\\.io/mode}").Output()
if err != nil {
if strings.Contains(modeInSecretAnnotation, "NotFound") {
if iaasPlatform != "aws" && iaasPlatform != "azure" && iaasPlatform != "gcp" {
mode = "passthrough"
return mode, nil
}
mode = "credsremoved"
return mode, nil
}
return "", err
}
if modeInSecretAnnotation == "insufficient" {
mode = "degraded"
return mode, nil
}
mode = modeInSecretAnnotation
return mode, nil
}
if iaasPlatform == "aws" {
if exutil.IsSTSCluster(oc) {
mode = "manualpodidentity"
return mode, nil
}
}
if iaasPlatform == "azure" {
if exutil.IsWorkloadIdentityCluster(oc) {
mode = "manualpodidentity"
return mode, nil
}
}
mode = "manual"
return mode, nil
}
func getRootSecretName(oc *exutil.CLI) (string, error) {
var rootSecretName string
iaasPlatform, err := getIaasPlatform(oc)
if err != nil {
return "", err
}
switch iaasPlatform {
case "aws":
rootSecretName = "aws-creds"
case "gcp":
rootSecretName = "gcp-credentials"
case "azure":
rootSecretName = "azure-credentials"
case "vsphere":
rootSecretName = "vsphere-creds"
case "openstack":
rootSecretName = "openstack-credentials"
case "ovirt":
rootSecretName = "ovirt-credentials"
default:
e2e.Logf("Unsupported platform: %v", iaasPlatform)
return "", nil
}
return rootSecretName, nil
}
func getIaasPlatform(oc *exutil.CLI) (string, error) {
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.status.platformStatus.type}").Output()
if err != nil {
return "", err
}
iaasPlatform := strings.ToLower(output)
if iaasPlatform == "external" {
externalPlatformNameOutput, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.spec.platformSpec.external.platformName}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
iaasPlatform = strings.ToLower(externalPlatformNameOutput)
}
return iaasPlatform, nil
}
func checkModeInMetric(oc *exutil.CLI, token string, mode string) error {
var (
data prometheusQueryResult
modeInMetric string
)
return wait.Poll(10*time.Second, 3*time.Minute, func() (bool, error) {
msg, _, err := oc.AsAdmin().WithoutNamespace().Run("exec").Args("-n", "openshift-monitoring", "prometheus-k8s-0", "-c", "prometheus", "--", "curl", "-k", "-H", fmt.Sprintf("Authorization: Bearer %v", token), "https://prometheus-k8s.openshift-monitoring.svc:9091/api/v1/query?query=cco_credentials_mode").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(msg).NotTo(o.BeEmpty())
json.Unmarshal([]byte(msg), &data)
modeInMetric = data.Data.Result[0].Metric.Mode
e2e.Logf("cco mode in metric is %v", modeInMetric)
if modeInMetric != mode {
e2e.Logf("cco mode should be %v, but is %v in metric", mode, modeInMetric)
return false, nil
}
return true, nil
})
}
func checkSTSStyle(oc *exutil.CLI, mode string) bool {
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "cloud-credentials", "-n", "openshift-ingress-operator", "-o=jsonpath={.data.credentials}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).NotTo(o.BeEmpty())
credentials, _ := base64.StdEncoding.DecodeString(output)
credConfig := strings.Split(string(credentials), "\n")
//Credentials items are in different order for self-managed OCP and ROSA, so sort firstly
sort.SliceStable(credConfig, func(i, j int) bool {
return strings.Compare(credConfig[i], credConfig[j]) < 0
})
if mode == "manualpodidentity" {
return strings.Contains(credConfig[0], "[default]") && strings.Contains(credConfig[1], "role_arn") && strings.Contains(credConfig[2], "sts_regional_endpoints") && strings.Contains(credConfig[3], "web_identity_token_file")
}
return strings.Contains(credConfig[0], "[default]") && strings.Contains(credConfig[1], "aws_access_key_id") && strings.Contains(credConfig[2], "aws_secret_access_key")
}
func patchResourceAsAdmin(oc *exutil.CLI, ns, resource, rsname, patch string) {
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args(resource, rsname, "--type=json", "-p", patch, "-n", ns).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
}
func (cr *credentialsRequest) create(oc *exutil.CLI) {
exutil.ApplyNsResourceFromTemplate(oc, "openshift-cloud-credential-operator", "--ignore-unknown-parameters=true", "-f", cr.template, "-p", "NAME="+cr.name, "NAMESPACE="+cr.namespace, "PROVIDER="+cr.provider)
}
// Check if CCO conditions are health
func checkCCOHealth(oc *exutil.CLI, mode string) {
availableStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Available")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(availableStatus).To(o.Equal("True"))
degradedStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(degradedStatus).To(o.Equal("False"))
progressingStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(progressingStatus).To(o.Equal("False"))
upgradeableStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Upgradeable")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
//when cco mode is manual or manual+sts, upgradeableStatus is "False" due to MissingUpgradeableAnnotation
if mode == "manual" || mode == "manualpodidentity" {
o.Expect(upgradeableStatus).To(o.Equal("False"))
upgradeableReason, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Upgradeable")].reason}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(upgradeableReason).To(o.Equal("MissingUpgradeableAnnotation"))
} else {
o.Expect(upgradeableStatus).To(o.Equal("True"))
}
}
// check webhook pod securityContext
func checkWebhookSecurityContext(oc *exutil.CLI, podnum int) {
webHookPodName := make([]string, podnum)
for i := 0; i < len(webHookPodName); i++ {
var err error
webHookPod, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", "-l", "app=pod-identity-webhook", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.items[*].metadata.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
webHookPodName = strings.Split(strings.TrimSpace(webHookPod), " ")
o.Expect(len(webHookPodName)).To(o.BeNumerically(">", 0))
e2e.Logf("webHookPodName is %s ", webHookPodName[i])
allowPrivilegeEscalation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.allowPrivilegeEscalation}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(allowPrivilegeEscalation).To(o.Equal("false"))
drop, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.capabilities.drop}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
dropAllCount := strings.Count(drop, "ALL")
o.Expect(dropAllCount).To(o.Equal(1))
runAsNonRoot, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.runAsNonRoot}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(runAsNonRoot).To(o.Equal("true"))
seccompProfileType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.seccompProfile.type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(seccompProfileType).To(o.Equal("RuntimeDefault"))
}
}
| package cloudcredential | ||||
function | openshift/openshift-tests-private | 83127739-15d9-4c57-9aa3-1b2121c226d7 | doOcpReq | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func doOcpReq(oc *exutil.CLI, verb OcpClientVerb, notEmpty bool, args ...string) string {
e2e.Logf("running command : oc %s %s", string(verb), strings.Join(args, " "))
res, err := oc.AsAdmin().WithoutNamespace().Run(string(verb)).Args(args...).Output()
o.Expect(err).ShouldNot(o.HaveOccurred())
if notEmpty {
o.Expect(res).ShouldNot(o.BeEmpty())
}
return res
} | cloudcredential | ||||
function | openshift/openshift-tests-private | e8044dff-1212-45fd-bfad-ab483ebbf0bd | getCloudCredentialMode | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func getCloudCredentialMode(oc *exutil.CLI) (string, error) {
var (
mode string
iaasPlatform string
rootSecretName string
err error
)
iaasPlatform, err = getIaasPlatform(oc)
if err != nil {
return "", err
}
if iaasPlatform == "none" || iaasPlatform == "baremetal" {
mode = "none" //mode none is for baremetal
return mode, nil
}
//Check if the cloud providers which support Manual mode only
if iaasPlatform == "ibmcloud" || iaasPlatform == "alibabacloud" || iaasPlatform == "nutanix" {
mode = "manual"
return mode, nil
}
modeInCloudCredential, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("cloudcredential", "cluster", "-o=jsonpath={.spec.credentialsMode}").Output()
if err != nil {
return "", err
}
if modeInCloudCredential != "Manual" {
rootSecretName, err = getRootSecretName(oc)
if err != nil {
return "", err
}
modeInSecretAnnotation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", rootSecretName, "-n=kube-system", "-o=jsonpath={.metadata.annotations.cloudcredential\\.openshift\\.io/mode}").Output()
if err != nil {
if strings.Contains(modeInSecretAnnotation, "NotFound") {
if iaasPlatform != "aws" && iaasPlatform != "azure" && iaasPlatform != "gcp" {
mode = "passthrough"
return mode, nil
}
mode = "credsremoved"
return mode, nil
}
return "", err
}
if modeInSecretAnnotation == "insufficient" {
mode = "degraded"
return mode, nil
}
mode = modeInSecretAnnotation
return mode, nil
}
if iaasPlatform == "aws" {
if exutil.IsSTSCluster(oc) {
mode = "manualpodidentity"
return mode, nil
}
}
if iaasPlatform == "azure" {
if exutil.IsWorkloadIdentityCluster(oc) {
mode = "manualpodidentity"
return mode, nil
}
}
mode = "manual"
return mode, nil
} | cloudcredential | ||||
function | openshift/openshift-tests-private | 1e44da36-a9e1-4bfa-97eb-2bf2339bf741 | getRootSecretName | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func getRootSecretName(oc *exutil.CLI) (string, error) {
var rootSecretName string
iaasPlatform, err := getIaasPlatform(oc)
if err != nil {
return "", err
}
switch iaasPlatform {
case "aws":
rootSecretName = "aws-creds"
case "gcp":
rootSecretName = "gcp-credentials"
case "azure":
rootSecretName = "azure-credentials"
case "vsphere":
rootSecretName = "vsphere-creds"
case "openstack":
rootSecretName = "openstack-credentials"
case "ovirt":
rootSecretName = "ovirt-credentials"
default:
e2e.Logf("Unsupported platform: %v", iaasPlatform)
return "", nil
}
return rootSecretName, nil
} | cloudcredential | |||||
function | openshift/openshift-tests-private | 72328533-23f5-4785-bc3d-2c8bff3d95ef | getIaasPlatform | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func getIaasPlatform(oc *exutil.CLI) (string, error) {
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.status.platformStatus.type}").Output()
if err != nil {
return "", err
}
iaasPlatform := strings.ToLower(output)
if iaasPlatform == "external" {
externalPlatformNameOutput, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.spec.platformSpec.external.platformName}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
iaasPlatform = strings.ToLower(externalPlatformNameOutput)
}
return iaasPlatform, nil
} | cloudcredential | ||||
function | openshift/openshift-tests-private | eb99a25e-04d9-457a-9333-6f41f6cf3768 | checkModeInMetric | ['"encoding/json"', '"fmt"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | ['prometheusQueryResult'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func checkModeInMetric(oc *exutil.CLI, token string, mode string) error {
var (
data prometheusQueryResult
modeInMetric string
)
return wait.Poll(10*time.Second, 3*time.Minute, func() (bool, error) {
msg, _, err := oc.AsAdmin().WithoutNamespace().Run("exec").Args("-n", "openshift-monitoring", "prometheus-k8s-0", "-c", "prometheus", "--", "curl", "-k", "-H", fmt.Sprintf("Authorization: Bearer %v", token), "https://prometheus-k8s.openshift-monitoring.svc:9091/api/v1/query?query=cco_credentials_mode").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(msg).NotTo(o.BeEmpty())
json.Unmarshal([]byte(msg), &data)
modeInMetric = data.Data.Result[0].Metric.Mode
e2e.Logf("cco mode in metric is %v", modeInMetric)
if modeInMetric != mode {
e2e.Logf("cco mode should be %v, but is %v in metric", mode, modeInMetric)
return false, nil
}
return true, nil
})
} | cloudcredential | |||
function | openshift/openshift-tests-private | 302fe9da-e977-4e0b-ab95-d43d19ff4017 | checkSTSStyle | ['"encoding/base64"', '"sort"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func checkSTSStyle(oc *exutil.CLI, mode string) bool {
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "cloud-credentials", "-n", "openshift-ingress-operator", "-o=jsonpath={.data.credentials}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).NotTo(o.BeEmpty())
credentials, _ := base64.StdEncoding.DecodeString(output)
credConfig := strings.Split(string(credentials), "\n")
//Credentials items are in different order for self-managed OCP and ROSA, so sort firstly
sort.SliceStable(credConfig, func(i, j int) bool {
return strings.Compare(credConfig[i], credConfig[j]) < 0
})
if mode == "manualpodidentity" {
return strings.Contains(credConfig[0], "[default]") && strings.Contains(credConfig[1], "role_arn") && strings.Contains(credConfig[2], "sts_regional_endpoints") && strings.Contains(credConfig[3], "web_identity_token_file")
}
return strings.Contains(credConfig[0], "[default]") && strings.Contains(credConfig[1], "aws_access_key_id") && strings.Contains(credConfig[2], "aws_secret_access_key")
} | cloudcredential | ||||
function | openshift/openshift-tests-private | fefc20f1-7a50-4d96-82b3-e2411ddd2237 | patchResourceAsAdmin | ['"encoding/json"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func patchResourceAsAdmin(oc *exutil.CLI, ns, resource, rsname, patch string) {
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args(resource, rsname, "--type=json", "-p", patch, "-n", ns).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
} | cloudcredential | ||||
function | openshift/openshift-tests-private | 5a3ba12b-c587-44fb-807d-03b6be59f1cb | create | ['credentialsRequest'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func (cr *credentialsRequest) create(oc *exutil.CLI) {
exutil.ApplyNsResourceFromTemplate(oc, "openshift-cloud-credential-operator", "--ignore-unknown-parameters=true", "-f", cr.template, "-p", "NAME="+cr.name, "NAMESPACE="+cr.namespace, "PROVIDER="+cr.provider)
} | cloudcredential | ||||
function | openshift/openshift-tests-private | 23901af1-1ee5-4266-8158-83dbcbc48b4d | checkCCOHealth | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func checkCCOHealth(oc *exutil.CLI, mode string) {
availableStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Available")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(availableStatus).To(o.Equal("True"))
degradedStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(degradedStatus).To(o.Equal("False"))
progressingStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(progressingStatus).To(o.Equal("False"))
upgradeableStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Upgradeable")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
//when cco mode is manual or manual+sts, upgradeableStatus is "False" due to MissingUpgradeableAnnotation
if mode == "manual" || mode == "manualpodidentity" {
o.Expect(upgradeableStatus).To(o.Equal("False"))
upgradeableReason, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Upgradeable")].reason}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(upgradeableReason).To(o.Equal("MissingUpgradeableAnnotation"))
} else {
o.Expect(upgradeableStatus).To(o.Equal("True"))
}
} | cloudcredential | |||||
function | openshift/openshift-tests-private | 26593b7f-461c-42d6-9926-cf123d763619 | checkWebhookSecurityContext | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential_util.go | func checkWebhookSecurityContext(oc *exutil.CLI, podnum int) {
webHookPodName := make([]string, podnum)
for i := 0; i < len(webHookPodName); i++ {
var err error
webHookPod, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", "-l", "app=pod-identity-webhook", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.items[*].metadata.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
webHookPodName = strings.Split(strings.TrimSpace(webHookPod), " ")
o.Expect(len(webHookPodName)).To(o.BeNumerically(">", 0))
e2e.Logf("webHookPodName is %s ", webHookPodName[i])
allowPrivilegeEscalation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.allowPrivilegeEscalation}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(allowPrivilegeEscalation).To(o.Equal("false"))
drop, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.capabilities.drop}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
dropAllCount := strings.Count(drop, "ALL")
o.Expect(dropAllCount).To(o.Equal(1))
runAsNonRoot, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.runAsNonRoot}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(runAsNonRoot).To(o.Equal("true"))
seccompProfileType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", webHookPodName[i], "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.seccompProfile.type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(seccompProfileType).To(o.Equal("RuntimeDefault"))
}
} | cloudcredential | ||||
test | openshift/openshift-tests-private | 70d512bb-b2a0-451b-a5be-f77f0dc8cc2f | cloudcredential | import (
"context"
"encoding/base64"
"fmt"
"io"
"io/ioutil"
"net/http"
"os"
"os/exec"
"path/filepath"
"regexp"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
"gopkg.in/yaml.v3"
"github.com/google/go-github/v57/github"
"github.com/tidwall/gjson"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"k8s.io/apimachinery/pkg/util/wait"
e2e "k8s.io/kubernetes/test/e2e/framework"
) | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | package cloudcredential
import (
"context"
"encoding/base64"
"fmt"
"io"
"io/ioutil"
"net/http"
"os"
"os/exec"
"path/filepath"
"regexp"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
"gopkg.in/yaml.v3"
"github.com/google/go-github/v57/github"
"github.com/tidwall/gjson"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"k8s.io/apimachinery/pkg/util/wait"
e2e "k8s.io/kubernetes/test/e2e/framework"
)
var _ = g.Describe("[sig-cco] Cluster_Operator CCO is enabled", func() {
defer g.GinkgoRecover()
var (
oc = exutil.NewCLI("default-cco", exutil.KubeConfigPath())
modeInMetric string
)
g.BeforeEach(func() {
exutil.SkipNoCapabilities(oc, ccoCap)
})
g.It("NonHyperShiftHOST-PstChkUpgrade-NonPreRelease-Author:mihuang-High-23352-Cloud credential operator resets progressing transition timestamp when it upgrades", func() {
exutil.By("Check if ns-23352 namespace exists")
ns := "ns-23352"
err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ns", ns).Execute()
if err != nil {
g.Skip("Skip the PstChkUpgrade test as ns-23352 namespace does not exist, PreChkUpgrade test did not run")
}
defer doOcpReq(oc, "delete", true, "ns", ns, "--ignore-not-found=true")
exutil.By("Get the progressingTransitionTimestamp before upgrade")
progressingTransitionTimestampCM := doOcpReq(oc, "get", true, "cm", "cm-23352", "-n", "ns-23352", `-o=jsonpath={.data.progressingTransitionTimestamp}`)
e2e.Logf("progressingTransitionTimestampCM: %s", progressingTransitionTimestampCM)
exutil.By("Check the progressing transition timestamp should be reset after upgrade")
progressingTransitionTimestampAfterUpgrade, err := time.Parse(time.RFC3339, doOcpReq(oc, "get", true, "clusteroperator", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("progressingTransitionTimestampAfterUpgrade: %s", progressingTransitionTimestampAfterUpgrade)
o.Expect(fmt.Sprintf("%s", progressingTransitionTimestampAfterUpgrade)).NotTo(o.Equal(progressingTransitionTimestampCM))
})
// author: [email protected]
// It is destructive case, will remove root credentials, so adding [Disruptive]. The case duration is greater than 5 minutes
// so adding [Slow]
g.It("NonHyperShiftHOST-Author:lwan-High-31768-Report the mode of cloud-credential operation as a metric [Slow][Disruptive]", func() {
exutil.By("Get cco mode from Cluster Resource")
modeInCR, err := getCloudCredentialMode(oc)
o.Expect(err).NotTo(o.HaveOccurred())
if modeInCR == "" {
e2e.Failf("Failed to get cco mode from Cluster Resource")
} else {
exutil.By("Check if cco mode in metric is the same as cco mode in cluster resources")
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
exutil.By("Check if cco mode in Metric is correct")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
if modeInCR == "mint" {
exutil.By("if cco is in mint mode currently, then run the below test")
exutil.By("Check cco mode when cco is in Passthrough mode")
//Force cco mode to Passthrough is NOT supported officially but is good for coverage on AWS/GCP Passthrough mode
e2e.Logf("Force cco mode to Passthrough")
originCCOMode, err := oc.AsAdmin().Run("get").Args("cloudcredential/cluster", "-o=jsonpath={.spec.credentialsMode}").Output()
if originCCOMode == "" {
originCCOMode = "\"\""
}
patchYaml := `
spec:
credentialsMode: ` + originCCOMode
err = oc.AsAdmin().Run("patch").Args("cloudcredential/cluster", "-p", `{"spec":{"credentialsMode":"Passthrough"}}`, "--type=merge").Execute()
defer func() {
err := oc.AsAdmin().Run("patch").Args("cloudcredential/cluster", "-p", patchYaml, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from cluster CR")
modeInCR, err = getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if cco mode in Metric is correct")
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
exutil.By("Check cco mode when root credential is removed when cco is not in manual mode")
e2e.Logf("remove root creds")
rootSecretName, err := getRootSecretName(oc)
o.Expect(err).NotTo(o.HaveOccurred())
rootSecretYaml, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", rootSecretName, "-n=kube-system", "-o=yaml").OutputToFile("root-secret.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", rootSecretName, "-n=kube-system").Execute()
defer func() {
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", rootSecretYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from cluster CR")
modeInCR, err = getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from Metric")
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
}
}
})
g.It("NonHyperShiftHOST-Author:mihuang-LEVEL0-Critical-33204-[cco-passthrough]IPI on azure with cco passthrough mode", func() {
exutil.By("Check if it's an azure cluster")
exutil.SkipIfPlatformTypeNot(oc, "azure")
mode, _ := getCloudCredentialMode(oc)
if mode != "passthrough" {
g.Skip("The cco mode is not passthrough - skipping test ...")
}
exutil.By("Check root credential has passthrough annotations")
o.Expect(doOcpReq(oc, "get", true, "secret", "-n", "kube-system", "azure-credentials", "-o=jsonpath={.metadata.annotations.cloudcredential\\.openshift\\.io/mode}")).Should(o.Equal("passthrough"))
})
//For bug https://bugzilla.redhat.com/show_bug.cgi?id=1940142
//For bug https://bugzilla.redhat.com/show_bug.cgi?id=1952891
g.It("NonHyperShiftHOST-Author:lwan-High-45415-[Bug 1940142] Reset CACert to correct path [Disruptive]", func() {
exutil.By("Check if it's an osp cluster")
exutil.SkipIfPlatformTypeNot(oc, "openstack")
exutil.By("Get openstack root credential clouds.yaml field")
goodCreds, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
goodCredsYaml := `
data:
clouds.yaml: ` + goodCreds
exutil.By("Check cacert path is correct")
CredsTXT, err := base64.StdEncoding.DecodeString(goodCreds)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if it's a kuryr cluster")
if !strings.Contains(string(CredsTXT), "cacert") {
g.Skip("Skip for non-kuryr cluster!")
}
o.Expect(CredsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
exutil.By("Patch cacert path to an wrong path")
var filename = "creds_45415.txt"
err = ioutil.WriteFile(filename, []byte(CredsTXT), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
wrongPath, err := exec.Command("bash", "-c", fmt.Sprintf("sed -i -e \"s/cacert: .*/cacert: path-no-exist/g\" %s && cat %s", filename, filename)).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(wrongPath).To(o.ContainSubstring("cacert: path-no-exist"))
o.Expect(wrongPath).NotTo(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
badCreds := base64.StdEncoding.EncodeToString(wrongPath)
wrongCredsYaml := `
data:
clouds.yaml: ` + badCreds
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", wrongCredsYaml).Execute()
defer func() {
oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", goodCredsYaml).Execute()
exutil.By("Wait for the storage operator to recover")
err = wait.Poll(10*time.Second, 60*time.Second, func() (bool, error) {
output, err := oc.AsAdmin().Run("get").Args("co", "storage").Output()
if err != nil {
e2e.Logf("Fail to get clusteroperator storage, error: %s. Trying again", err)
return false, nil
}
if matched, _ := regexp.MatchString("True.*False.*False", output); matched {
e2e.Logf("clusteroperator storage is recover to normal:\n%s", output)
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "clusteroperator storage is not recovered to normal")
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check cco change wrong path to correct one")
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err := base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
o.Expect(credsTXT).NotTo(o.ContainSubstring("cacert: path-no-exist"))
exutil.By("Patch cacert path to an empty path")
wrongPath, err = exec.Command("bash", "-c", fmt.Sprintf("sed -i -e \"s/cacert: .*/cacert:/g\" %s && cat %s", filename, filename)).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(wrongPath).To(o.ContainSubstring("cacert:"))
o.Expect(wrongPath).NotTo(o.ContainSubstring("cacert: path-no-exist"))
badCreds = base64.StdEncoding.EncodeToString(wrongPath)
wrongCredsYaml = `
data:
clouds.yaml: ` + badCreds
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", wrongCredsYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check cco remove cacert field when it's value is empty")
output, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err = base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).NotTo(o.ContainSubstring("cacert:"))
exutil.By("recover root credential")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", goodCredsYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
output, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err = base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
})
g.It("NonHyperShiftHOST-ROSA-OSD_CCS-Author:jshu-LEVEL0-Critical-36498-CCO credentials secret change to STS-style", func() {
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform != "aws" {
g.Skip("IAAS platform is " + iaasPlatform + " while 36498 is for AWS - skipping test ...")
}
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster is %v", mode)
o.Expect(err).NotTo(o.HaveOccurred())
if mode == "manual" {
g.Skip(" Test case 36498 is not for cco mode=manual - skipping test ...")
}
if !checkSTSStyle(oc, mode) {
g.Fail("The secret format didn't pass STS style check.")
}
})
// Author: [email protected] [email protected]
g.It("Author:jshu-NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Medium-50869-High-53283-High-77285- CCO Pod Security Admission change", func() {
exutil.By("1.Check cloud-credential-operator pod")
ccoPodName, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", "-l", "app=cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.items[*].metadata.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
allowPrivilegeEscalation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.allowPrivilegeEscalation}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(allowPrivilegeEscalation).NotTo(o.ContainSubstring("true"))
drop, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.capabilities.drop}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
dropAllCount := strings.Count(drop, "ALL")
o.Expect(dropAllCount).To(o.Equal(2))
runAsNonRoot, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.runAsNonRoot}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(runAsNonRoot).To(o.Equal("true"))
seccompProfileType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.seccompProfile.type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(seccompProfileType).To(o.Equal("RuntimeDefault"))
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform == "aws" || iaasPlatform == "azure" || iaasPlatform == "gcp" {
exutil.By(fmt.Sprintf("2.Check pod-identity-webhook pod for %s", iaasPlatform))
if exutil.IsSNOCluster(oc) {
checkWebhookSecurityContext(oc, 1)
} else {
checkWebhookSecurityContext(oc, 2)
}
}
})
g.It("NonHyperShiftHOST-Author:jshu-Medium-48360-Reconciliation of aws pod identity mutating webhook did not happen [Disruptive]", func() {
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform != "aws" {
g.Skip("IAAS platform is " + iaasPlatform + " while 48360 is for AWS - skipping test ...")
}
exutil.By("1.Check the Mutating Webhook Configuration service port is 443")
port, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(port).To(o.Equal("443"))
exutil.By("2.Scale down cco pod")
output, err := oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "--replicas=0").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("scaled"))
exutil.By("3.Change the Mutating Webhook Configuration port to 444")
patchContent := "[{\"op\": \"replace\", \"path\": \"/webhooks/0/clientConfig/service/port\", \"value\":444}]"
patchResourceAsAdmin(oc, oc.Namespace(), "mutatingwebhookconfiguration", "pod-identity-webhook", patchContent)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("4.Now the Mutating Webhook Configuration service port is 444")
port, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(port).To(o.Equal("444"))
exutil.By("5.1.Scale up cco pod")
output, err = oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "--replicas=1").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("scaled"))
//Need wait for some time to verify if the port reset to 443
exutil.By("5.2.Check the Mutating Webhook Configuration service port is reset to 443")
errWait := wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
result, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
if err != nil || result != "443" {
e2e.Logf("Encountered error or the port is NOT reset yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(errWait, "The port is not reset to 443")
})
g.It("NonHyperShiftHOST-Author:jshu-Medium-45975-Test cco condition changes [Disruptive]", func() {
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("cco mode in cluster is %v", mode)
if mode == "manual" || mode == "manualpodidentity" {
g.Skip(" Test case 45975 is not for cco mode manual - skipping test ...")
}
//Check IAAS platform type
iaasPlatform, err := getIaasPlatform(oc)
o.Expect(err).NotTo(o.HaveOccurred())
var providerSpec string
switch iaasPlatform {
case "aws":
providerSpec = "AWSProviderSpec"
case "azure":
providerSpec = "AzureProviderSpec"
case "gcp":
providerSpec = "GCPProviderSpec"
case "openstack":
providerSpec = "OpenStackProviderSpec"
case "vsphere":
providerSpec = "VSphereProviderSpec"
default:
g.Skip("IAAS platform is " + iaasPlatform + " which is NOT supported by 45975 - skipping test ...")
}
exutil.By("Degraded condition status is False at first")
degradedStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(degradedStatus).To(o.Equal("False"))
exutil.By("Create 1st CredentialsRequest whose namespace does not exist")
testDataDir := exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
crTemp := filepath.Join(testDataDir, "credentials_request.yaml")
crName1 := "cloud-credential-operator-iam-ro-1"
crNamespace := "namespace-does-not-exist"
credentialsRequest1 := credentialsRequest{
name: crName1,
namespace: crNamespace,
provider: providerSpec,
template: crTemp,
}
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName1, "-n", "openshift-cloud-credential-operator", "--ignore-not-found").Execute()
credentialsRequest1.create(oc)
exutil.By("Check the Degraded status is True now and save the timestamp")
err = wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
degradedStatus, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
if err != nil || degradedStatus != "True" {
e2e.Logf("Degraded status is NOT True yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(err, "Degraded status is NOT set to True due to wrong CR.")
//save lastTransitionTime of Degraded condition
oldDegradedTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
//save lastTransitionTime of Progressing condition
oldProgressingTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create 2nd CredentialsRequest whose namespace does not exist")
crName2 := "cloud-credential-operator-iam-ro-2"
credentialsRequest2 := credentialsRequest{
name: crName2,
namespace: crNamespace,
provider: providerSpec,
template: crTemp,
}
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName2, "-n", "openshift-cloud-credential-operator", "--ignore-not-found").Execute()
credentialsRequest2.create(oc)
exutil.By("Check 2 CR reporting errors and lastTransitionTime of Degraded and Progressing not changed")
err = wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
progressingMessage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].message}`).Output()
if err != nil || !strings.Contains(progressingMessage, "2 reporting errors") {
e2e.Logf("CCO didn't detect 2nd wrong CR yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(err, "CCO didn't detect 2nd wrong CR finally.")
//compare the lastTransitionTime
newDegradedTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(newDegradedTimestamp).To(o.Equal(oldDegradedTimestamp))
newProgressingTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(newProgressingTimestamp).To(o.Equal(oldProgressingTimestamp))
})
//For bug https://bugzilla.redhat.com/show_bug.cgi?id=1977319
g.It("NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-High-45219-A fresh cluster should not have stale CR", func() {
output, _ := oc.AsAdmin().WithoutNamespace().Run("get").Args("service", "controller-manager-service", "-n", "openshift-cloud-credential-operator").Output()
o.Expect(output).To(o.ContainSubstring("Error from server (NotFound)"))
})
g.It("NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-Critical-34470-Cloud credential operator health check", func() {
exutil.By("Check CCO status conditions")
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster is %v", mode)
o.Expect(err).NotTo(o.HaveOccurred())
checkCCOHealth(oc, mode)
exutil.By("Check CCO imagePullPolicy configuration")
imagePullPolicy, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.template.spec.containers[1].imagePullPolicy}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(imagePullPolicy).To(o.Equal("IfNotPresent"))
})
g.It("NonHyperShiftHOST-OSD_CCS-ARO-Author:mihuang-LEVEL0-Critical-66538-Azure workload identity cluster healthy check.", func() {
mode, _ := getCloudCredentialMode(oc)
if !(exutil.CheckPlatform(oc) == "azure" && mode == "manualpodidentity") {
g.Skip("The cluster is not Azure Workload Identity Cluster - skipping test ...")
}
exutil.By("Check CCO status conditions")
checkCCOHealth(oc, mode)
exutil.By("The Azure workload identity cluster does not have `root` credentials")
cmdOut, err := oc.AsAdmin().Run("get").Args("secret", "azure-credentials", "-n", "kube-system").Output()
o.Expect(err).Should(o.HaveOccurred())
o.Expect(cmdOut).To(o.ContainSubstring("Error from server (NotFound)"))
exutil.By("The secret should contain azure_federated_token_file instead of azure credential keys.")
o.Expect(strings.Contains(doOcpReq(oc, "get", true, "secrets", "-n", "openshift-image-registry", "installer-cloud-credentials", "-o=jsonpath={.data}"), "azure_federated_token_file")).Should(o.BeTrue())
})
g.It("Author:fxie-Critical-64885-CCO-based flow for olm managed operators and AWS STS", func() {
exutil.SkipIfPlatformTypeNot(oc, "aws")
if !exutil.IsSTSCluster(oc) {
g.Skip("This test case is AWS STS only, skipping")
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "64885"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
stsIAMRoleARN = "whatever"
cloudTokenPath = "anything"
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "AWSProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Inserting an stsIAMRoleARN to the CR")
stsIAMRoleARNPatch := `
spec:
providerSpec:
stsIAMRoleARN: ` + stsIAMRoleARN
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", stsIAMRoleARNPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
stdout, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("extract").
Args("Secret/"+targetSecretName, "-n", targetNs, "--keys", "credentials", "--to", "-").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// The Secret does not contain any sensitive info
e2e.Logf("Secret extracted = %v", stdout)
o.Expect(stdout).To(o.ContainSubstring("[default]"))
o.Expect(stdout).To(o.ContainSubstring("sts_regional_endpoints = regional"))
o.Expect(stdout).To(o.ContainSubstring("role_arn = " + stsIAMRoleARN))
o.Expect(stdout).To(o.ContainSubstring("web_identity_token_file = " + defaultSTSCloudTokenPath))
exutil.By("Inserting a cloudTokenPath to the CR")
cloudTokenPathPatch := `
spec:
cloudTokenPath: ` + cloudTokenPath
err = oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", cloudTokenPathPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is updated in the correct way")
o.Eventually(func() bool {
stdout, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("extract").
Args("Secret/"+targetSecretName, "-n", targetNs, "--keys", "credentials", "--to", "-").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// The Secret does not contain any sensitive info
e2e.Logf("Secret extracted = %v", stdout)
return strings.Contains(stdout, "web_identity_token_file = "+cloudTokenPath)
}).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
})
g.It("NonHyperShiftHOST-OSD_CCS-ARO-Author:jshu-Critical-69971-Azure workload identity management for olm managed operators", func() {
exutil.SkipIfPlatformTypeNot(oc, "azure")
if !exutil.IsWorkloadIdentityCluster(oc) {
g.Skip("This test case is for Azure Workload Identity only, skipping")
}
//Provide the following Azure Credentials with fake values
azureCredList := []azureCredential{
{
key: "azure_subscription_id",
value: "12345678-1234-1234-1234-123456789ab",
},
{
key: "azure_tenant_id",
value: "23456789-2345-2345-2345-23456789abcd",
},
{
key: "azure_region",
value: "eastus",
},
{
key: "azure_client_id",
value: "3456789a-3456-3456-3456-23456789abcde",
},
{
key: "azure_federated_token_file",
value: "/var/run/secrets/token",
},
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "69971"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating the dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "AzureProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Patching the Azure Credentials and cloudTokenPath to the CR")
crPatch := `
spec:
cloudTokenPath: ` + azureCredList[4].value + `
providerSpec:
azureSubscriptionID: ` + azureCredList[0].value + `
azureTenantID: ` + azureCredList[1].value + `
azureRegion: ` + azureCredList[2].value + `
azureClientID: ` + azureCredList[3].value
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", crPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
for _, azureCred := range azureCredList {
credential, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/"+targetSecretName, "-n", targetNs, "--keys", azureCred.key, "--to", "-").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credential).To(o.ContainSubstring(azureCred.value))
}
})
g.It("Author:jshu-NonHyperShiftHOST-OSD_CCS-Critical-75429-GCP workload identity management for olm managed operators", func() {
exutil.SkipIfPlatformTypeNot(oc, "gcp")
if !exutil.IsWorkloadIdentityCluster(oc) {
g.Skip("This test case is for GCP Workload Identity only, skipping")
}
//Provide the following GCP Credentials with fake values
gcpCredList := []gcpCredential{
{
key: "audience",
value: "//iam.googleapis.com/projects/1042363005003/locations/global/workloadIdentityPools/cco-test/providers/cco-test",
},
{
key: "serviceAccountEmail",
value: "[email protected]",
},
{
key: "cloudTokenPath",
value: "/var/run/secrets/token",
},
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "75429"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating the dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "GCPProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Patching the GCP Credentials and cloudTokenPath to the CR")
crPatch := `
spec:
cloudTokenPath: ` + gcpCredList[2].value + `
providerSpec:
audience: ` + gcpCredList[0].value + `
serviceAccountEmail: ` + gcpCredList[1].value
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", crPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
credentialBase64, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", targetSecretName, "-n", targetNs, "-o=jsonpath={.data.service_account\\.json}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credential, err := base64.StdEncoding.DecodeString(credentialBase64)
o.Expect(err).NotTo(o.HaveOccurred())
//compare audience
gen_audience := gjson.Get(string(credential), `audience`).String()
o.Expect(gen_audience).To(o.Equal(gcpCredList[0].value))
//check serviceAccountEmail
gen_service_account := gjson.Get(string(credential), `service_account_impersonation_url`).String()
o.Expect(gen_service_account).To(o.ContainSubstring(gcpCredList[1].value))
//compare token path
gen_token_path := gjson.Get(string(credential), `credential_source.file`).String()
o.Expect(gen_token_path).To(o.Equal(gcpCredList[2].value))
})
})
var _ = g.Describe("[sig-cco] Cluster_Operator CCO is disabled", func() {
defer g.GinkgoRecover()
var (
oc = exutil.NewCLI("cco", exutil.KubeConfigPath())
)
g.BeforeEach(func() {
exutil.SkipIfCapEnabled(oc, ccoCap)
})
/*
Only deals with the first half of OCP-68220 (makes sure CCO-related resources are not found in-cluster).
The second half of OCP-68220 (day-2 enablement) will be covered by CI Profiles in Prow.
Hard-coding resources-to-check is avoided since:
- It leaves us a maintenance burden
- The test case will not be able to detect such scenario when a resource is added (but not annotated) in the future
*/
g.It("NonHyperShiftHOST-Author:fxie-Critical-68220-Leverage Composable OpenShift feature to make cloud-credential optional", func() {
var (
getManifestContent = func(manifest *github.RepositoryContent) []byte {
// Prefer manifest.Content
if content, _ := manifest.GetContent(); content != "" {
return []byte(content)
}
// Fall back to downloadURL
downloadURL := manifest.GetDownloadURL()
o.Expect(downloadURL).NotTo(o.BeEmpty())
req, err := http.NewRequest("GET", downloadURL, nil)
o.Expect(err).NotTo(o.HaveOccurred())
res, err := http.DefaultClient.Do(req)
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
_ = res.Body.Close()
}()
content, err := io.ReadAll(res.Body)
o.Expect(err).NotTo(o.HaveOccurred())
return content
}
/*
Here we avoid deserializing manifests through a runtime.Scheme since
installing all required types (to the scheme) would bring in quite a few
dependencies, making our test repo unnecessarily heavy.
*/
resourceInfoFromManifest = func(manifest []byte) (string, string, string) {
var deserializedManifest map[string]any
err := yaml.Unmarshal(manifest, &deserializedManifest)
o.Expect(err).NotTo(o.HaveOccurred())
groupVersion, ok := deserializedManifest["apiVersion"].(string)
o.Expect(ok).To(o.BeTrue())
groupVersionSlice := strings.Split(groupVersion, "/")
kind, ok := deserializedManifest["kind"].(string)
o.Expect(ok).To(o.BeTrue())
// The oc client is smart enough to map kind to resource before making an API call.
// There's no need to query the discovery endpoint of the API server ourselves to obtain the gvr.
var resourceType, namespace, name string
switch len(groupVersionSlice) {
// The resource is a part of the core group
case 1:
resourceType = kind
// The resource is not a part of the core group
case 2:
resourceType = fmt.Sprintf("%s.%s.%s", kind, groupVersionSlice[1], groupVersionSlice[0])
default:
e2e.Failf("Unexpected apiVersion format")
}
metadata, ok := deserializedManifest["metadata"].(map[string]any)
o.Expect(ok).To(o.BeTrue())
if _, isNamespaced := metadata["namespace"]; isNamespaced {
namespace, ok = metadata["namespace"].(string)
o.Expect(ok).To(o.BeTrue())
}
name, ok = metadata["name"].(string)
o.Expect(ok).To(o.BeTrue())
e2e.Logf("Resource type = %v, namespace = %v, name = %v", resourceType, namespace, name)
return resourceType, namespace, name
}
)
// Get GitHub client
ghClient := github.NewClient(nil)
// Authenticate for a much larger rate limit
if ghToken := os.Getenv("GITHUB_TOKEN"); ghToken != "" {
ghClient = ghClient.WithAuthToken(ghToken)
}
// Get cluster version
majorMinorVersion, _, err := exutil.GetClusterVersion(oc)
o.Expect(err).NotTo(o.HaveOccurred())
targetBranch := fmt.Sprintf("release-%s", majorMinorVersion)
// Get manifest dir of the CCO repo
// There's no need to verify the absence of CredentialsRequests defined in other repos.
// We only need to make sure the corresponding CRD is not found in-cluster.
e2e.Logf("Listing manifest directory of branch %v", targetBranch)
_, dir, _, err := ghClient.
Repositories.
GetContents(context.Background(), "openshift", ccoRepo, ccoManifestPath, &github.RepositoryContentGetOptions{
Ref: targetBranch,
})
o.Expect(err).NotTo(o.HaveOccurred())
// Iterate through items in the manifest dir
for _, manifest := range dir {
if manifest.GetType() != "file" {
continue
}
manifestName := manifest.GetName()
if !strings.HasSuffix(manifestName, ".yaml") && !strings.HasSuffix(manifestName, ".yml") {
continue
}
e2e.Logf("Getting content of manifest %v", manifestName)
content := getManifestContent(manifest)
e2e.Logf("Extracting resource info from manifest")
resourceType, namespace, name := resourceInfoFromManifest(content)
e2e.Logf("Requesting manifest against the API server")
getReqArgs := []string{resourceType, name}
if namespace != "" {
getReqArgs = append(getReqArgs, "-n", namespace)
}
// err is the error returned by executing an exec.Command
// stderr captures the original error message return by the API server
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("get").Args(getReqArgs...).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.Or(
o.ContainSubstring("not found"),
o.ContainSubstring("the server doesn't have a resource type"),
))
}
})
})
| package cloudcredential | ||||
test case | openshift/openshift-tests-private | e848d206-f3d9-4194-82a2-3a7325b91335 | NonHyperShiftHOST-PstChkUpgrade-NonPreRelease-Author:mihuang-High-23352-Cloud credential operator resets progressing transition timestamp when it upgrades | ['"fmt"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-PstChkUpgrade-NonPreRelease-Author:mihuang-High-23352-Cloud credential operator resets progressing transition timestamp when it upgrades", func() {
exutil.By("Check if ns-23352 namespace exists")
ns := "ns-23352"
err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ns", ns).Execute()
if err != nil {
g.Skip("Skip the PstChkUpgrade test as ns-23352 namespace does not exist, PreChkUpgrade test did not run")
}
defer doOcpReq(oc, "delete", true, "ns", ns, "--ignore-not-found=true")
exutil.By("Get the progressingTransitionTimestamp before upgrade")
progressingTransitionTimestampCM := doOcpReq(oc, "get", true, "cm", "cm-23352", "-n", "ns-23352", `-o=jsonpath={.data.progressingTransitionTimestamp}`)
e2e.Logf("progressingTransitionTimestampCM: %s", progressingTransitionTimestampCM)
exutil.By("Check the progressing transition timestamp should be reset after upgrade")
progressingTransitionTimestampAfterUpgrade, err := time.Parse(time.RFC3339, doOcpReq(oc, "get", true, "clusteroperator", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("progressingTransitionTimestampAfterUpgrade: %s", progressingTransitionTimestampAfterUpgrade)
o.Expect(fmt.Sprintf("%s", progressingTransitionTimestampAfterUpgrade)).NotTo(o.Equal(progressingTransitionTimestampCM))
}) | |||||
test case | openshift/openshift-tests-private | 09376c68-8087-4bf1-b85f-7dab3fc83d30 | NonHyperShiftHOST-Author:lwan-High-31768-Report the mode of cloud-credential operation as a metric [Slow][Disruptive] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:lwan-High-31768-Report the mode of cloud-credential operation as a metric [Slow][Disruptive]", func() {
exutil.By("Get cco mode from Cluster Resource")
modeInCR, err := getCloudCredentialMode(oc)
o.Expect(err).NotTo(o.HaveOccurred())
if modeInCR == "" {
e2e.Failf("Failed to get cco mode from Cluster Resource")
} else {
exutil.By("Check if cco mode in metric is the same as cco mode in cluster resources")
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
exutil.By("Check if cco mode in Metric is correct")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
if modeInCR == "mint" {
exutil.By("if cco is in mint mode currently, then run the below test")
exutil.By("Check cco mode when cco is in Passthrough mode")
//Force cco mode to Passthrough is NOT supported officially but is good for coverage on AWS/GCP Passthrough mode
e2e.Logf("Force cco mode to Passthrough")
originCCOMode, err := oc.AsAdmin().Run("get").Args("cloudcredential/cluster", "-o=jsonpath={.spec.credentialsMode}").Output()
if originCCOMode == "" {
originCCOMode = "\"\""
}
patchYaml := `
spec:
credentialsMode: ` + originCCOMode
err = oc.AsAdmin().Run("patch").Args("cloudcredential/cluster", "-p", `{"spec":{"credentialsMode":"Passthrough"}}`, "--type=merge").Execute()
defer func() {
err := oc.AsAdmin().Run("patch").Args("cloudcredential/cluster", "-p", patchYaml, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from cluster CR")
modeInCR, err = getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if cco mode in Metric is correct")
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
exutil.By("Check cco mode when root credential is removed when cco is not in manual mode")
e2e.Logf("remove root creds")
rootSecretName, err := getRootSecretName(oc)
o.Expect(err).NotTo(o.HaveOccurred())
rootSecretYaml, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", rootSecretName, "-n=kube-system", "-o=yaml").OutputToFile("root-secret.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", rootSecretName, "-n=kube-system").Execute()
defer func() {
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", rootSecretYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from cluster CR")
modeInCR, err = getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster CR is %v", modeInCR)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Get cco mode from Metric")
err = checkModeInMetric(oc, token, modeInCR)
if err != nil {
e2e.Failf("Failed to check cco mode metric after waiting up to 3 minutes, cco mode should be %v, but is %v in metric", modeInCR, modeInMetric)
}
}
}
}) | ||||||
test case | openshift/openshift-tests-private | 2c3fb111-ada1-449d-8cd3-d940c9e372cb | NonHyperShiftHOST-Author:mihuang-LEVEL0-Critical-33204-[cco-passthrough]IPI on azure with cco passthrough mode | ['"io"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:mihuang-LEVEL0-Critical-33204-[cco-passthrough]IPI on azure with cco passthrough mode", func() {
exutil.By("Check if it's an azure cluster")
exutil.SkipIfPlatformTypeNot(oc, "azure")
mode, _ := getCloudCredentialMode(oc)
if mode != "passthrough" {
g.Skip("The cco mode is not passthrough - skipping test ...")
}
exutil.By("Check root credential has passthrough annotations")
o.Expect(doOcpReq(oc, "get", true, "secret", "-n", "kube-system", "azure-credentials", "-o=jsonpath={.metadata.annotations.cloudcredential\\.openshift\\.io/mode}")).Should(o.Equal("passthrough"))
}) | |||||
test case | openshift/openshift-tests-private | 9faa86cd-c429-4aaa-b521-6a1cea0ffb48 | NonHyperShiftHOST-Author:lwan-High-45415-[Bug 1940142] Reset CACert to correct path [Disruptive] | ['"encoding/base64"', '"fmt"', '"io/ioutil"', '"os"', '"os/exec"', '"regexp"', '"strings"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:lwan-High-45415-[Bug 1940142] Reset CACert to correct path [Disruptive]", func() {
exutil.By("Check if it's an osp cluster")
exutil.SkipIfPlatformTypeNot(oc, "openstack")
exutil.By("Get openstack root credential clouds.yaml field")
goodCreds, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
goodCredsYaml := `
data:
clouds.yaml: ` + goodCreds
exutil.By("Check cacert path is correct")
CredsTXT, err := base64.StdEncoding.DecodeString(goodCreds)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if it's a kuryr cluster")
if !strings.Contains(string(CredsTXT), "cacert") {
g.Skip("Skip for non-kuryr cluster!")
}
o.Expect(CredsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
exutil.By("Patch cacert path to an wrong path")
var filename = "creds_45415.txt"
err = ioutil.WriteFile(filename, []byte(CredsTXT), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
wrongPath, err := exec.Command("bash", "-c", fmt.Sprintf("sed -i -e \"s/cacert: .*/cacert: path-no-exist/g\" %s && cat %s", filename, filename)).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(wrongPath).To(o.ContainSubstring("cacert: path-no-exist"))
o.Expect(wrongPath).NotTo(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
badCreds := base64.StdEncoding.EncodeToString(wrongPath)
wrongCredsYaml := `
data:
clouds.yaml: ` + badCreds
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", wrongCredsYaml).Execute()
defer func() {
oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", goodCredsYaml).Execute()
exutil.By("Wait for the storage operator to recover")
err = wait.Poll(10*time.Second, 60*time.Second, func() (bool, error) {
output, err := oc.AsAdmin().Run("get").Args("co", "storage").Output()
if err != nil {
e2e.Logf("Fail to get clusteroperator storage, error: %s. Trying again", err)
return false, nil
}
if matched, _ := regexp.MatchString("True.*False.*False", output); matched {
e2e.Logf("clusteroperator storage is recover to normal:\n%s", output)
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "clusteroperator storage is not recovered to normal")
}()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check cco change wrong path to correct one")
output, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err := base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
o.Expect(credsTXT).NotTo(o.ContainSubstring("cacert: path-no-exist"))
exutil.By("Patch cacert path to an empty path")
wrongPath, err = exec.Command("bash", "-c", fmt.Sprintf("sed -i -e \"s/cacert: .*/cacert:/g\" %s && cat %s", filename, filename)).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(wrongPath).To(o.ContainSubstring("cacert:"))
o.Expect(wrongPath).NotTo(o.ContainSubstring("cacert: path-no-exist"))
badCreds = base64.StdEncoding.EncodeToString(wrongPath)
wrongCredsYaml = `
data:
clouds.yaml: ` + badCreds
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", wrongCredsYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check cco remove cacert field when it's value is empty")
output, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err = base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).NotTo(o.ContainSubstring("cacert:"))
exutil.By("recover root credential")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "openstack-credentials", "-n=kube-system", "--type=merge", "-p", goodCredsYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
output, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "openstack-credentials", "-n=kube-system", "-o=jsonpath={.data.clouds\\.yaml}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credsTXT, err = base64.StdEncoding.DecodeString(output)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credsTXT).To(o.ContainSubstring("cacert: /etc/kubernetes/static-pod-resources/configmaps/cloud-config/ca-bundle.pem"))
}) | |||||
test case | openshift/openshift-tests-private | 78305086-bc30-4a58-9a2f-5f28501f9bf0 | NonHyperShiftHOST-ROSA-OSD_CCS-Author:jshu-LEVEL0-Critical-36498-CCO credentials secret change to STS-style | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-ROSA-OSD_CCS-Author:jshu-LEVEL0-Critical-36498-CCO credentials secret change to STS-style", func() {
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform != "aws" {
g.Skip("IAAS platform is " + iaasPlatform + " while 36498 is for AWS - skipping test ...")
}
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster is %v", mode)
o.Expect(err).NotTo(o.HaveOccurred())
if mode == "manual" {
g.Skip(" Test case 36498 is not for cco mode=manual - skipping test ...")
}
if !checkSTSStyle(oc, mode) {
g.Fail("The secret format didn't pass STS style check.")
}
}) | ||||||
test case | openshift/openshift-tests-private | d4a233af-14d6-41ce-b40d-8ef8f53334df | Author:jshu-NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Medium-50869-High-53283-High-77285- CCO Pod Security Admission change | ['"fmt"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("Author:jshu-NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Medium-50869-High-53283-High-77285- CCO Pod Security Admission change", func() {
exutil.By("1.Check cloud-credential-operator pod")
ccoPodName, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", "-l", "app=cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.items[*].metadata.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
allowPrivilegeEscalation, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.allowPrivilegeEscalation}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(allowPrivilegeEscalation).NotTo(o.ContainSubstring("true"))
drop, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.containers[*].securityContext.capabilities.drop}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
dropAllCount := strings.Count(drop, "ALL")
o.Expect(dropAllCount).To(o.Equal(2))
runAsNonRoot, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.runAsNonRoot}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(runAsNonRoot).To(o.Equal("true"))
seccompProfileType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pod", ccoPodName, "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.securityContext.seccompProfile.type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(seccompProfileType).To(o.Equal("RuntimeDefault"))
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform == "aws" || iaasPlatform == "azure" || iaasPlatform == "gcp" {
exutil.By(fmt.Sprintf("2.Check pod-identity-webhook pod for %s", iaasPlatform))
if exutil.IsSNOCluster(oc) {
checkWebhookSecurityContext(oc, 1)
} else {
checkWebhookSecurityContext(oc, 2)
}
}
}) | |||||
test case | openshift/openshift-tests-private | 5dfcffcd-96ec-4b64-bc6c-0a455a93c90f | NonHyperShiftHOST-Author:jshu-Medium-48360-Reconciliation of aws pod identity mutating webhook did not happen [Disruptive] | ['"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:jshu-Medium-48360-Reconciliation of aws pod identity mutating webhook did not happen [Disruptive]", func() {
//Check IAAS platform type
iaasPlatform := exutil.CheckPlatform(oc)
if iaasPlatform != "aws" {
g.Skip("IAAS platform is " + iaasPlatform + " while 48360 is for AWS - skipping test ...")
}
exutil.By("1.Check the Mutating Webhook Configuration service port is 443")
port, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(port).To(o.Equal("443"))
exutil.By("2.Scale down cco pod")
output, err := oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "--replicas=0").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("scaled"))
exutil.By("3.Change the Mutating Webhook Configuration port to 444")
patchContent := "[{\"op\": \"replace\", \"path\": \"/webhooks/0/clientConfig/service/port\", \"value\":444}]"
patchResourceAsAdmin(oc, oc.Namespace(), "mutatingwebhookconfiguration", "pod-identity-webhook", patchContent)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("4.Now the Mutating Webhook Configuration service port is 444")
port, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(port).To(o.Equal("444"))
exutil.By("5.1.Scale up cco pod")
output, err = oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "--replicas=1").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("scaled"))
//Need wait for some time to verify if the port reset to 443
exutil.By("5.2.Check the Mutating Webhook Configuration service port is reset to 443")
errWait := wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
result, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("mutatingwebhookconfiguration", "pod-identity-webhook", "-o=jsonpath={.webhooks[].clientConfig.service.port}").Output()
if err != nil || result != "443" {
e2e.Logf("Encountered error or the port is NOT reset yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(errWait, "The port is not reset to 443")
}) | |||||
test case | openshift/openshift-tests-private | b9b7f175-8387-4b72-a9f2-48b321a17d55 | NonHyperShiftHOST-Author:jshu-Medium-45975-Test cco condition changes [Disruptive] | ['"path/filepath"', '"strings"', '"time"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:jshu-Medium-45975-Test cco condition changes [Disruptive]", func() {
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("cco mode in cluster is %v", mode)
if mode == "manual" || mode == "manualpodidentity" {
g.Skip(" Test case 45975 is not for cco mode manual - skipping test ...")
}
//Check IAAS platform type
iaasPlatform, err := getIaasPlatform(oc)
o.Expect(err).NotTo(o.HaveOccurred())
var providerSpec string
switch iaasPlatform {
case "aws":
providerSpec = "AWSProviderSpec"
case "azure":
providerSpec = "AzureProviderSpec"
case "gcp":
providerSpec = "GCPProviderSpec"
case "openstack":
providerSpec = "OpenStackProviderSpec"
case "vsphere":
providerSpec = "VSphereProviderSpec"
default:
g.Skip("IAAS platform is " + iaasPlatform + " which is NOT supported by 45975 - skipping test ...")
}
exutil.By("Degraded condition status is False at first")
degradedStatus, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(degradedStatus).To(o.Equal("False"))
exutil.By("Create 1st CredentialsRequest whose namespace does not exist")
testDataDir := exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
crTemp := filepath.Join(testDataDir, "credentials_request.yaml")
crName1 := "cloud-credential-operator-iam-ro-1"
crNamespace := "namespace-does-not-exist"
credentialsRequest1 := credentialsRequest{
name: crName1,
namespace: crNamespace,
provider: providerSpec,
template: crTemp,
}
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName1, "-n", "openshift-cloud-credential-operator", "--ignore-not-found").Execute()
credentialsRequest1.create(oc)
exutil.By("Check the Degraded status is True now and save the timestamp")
err = wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
degradedStatus, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].status}`).Output()
if err != nil || degradedStatus != "True" {
e2e.Logf("Degraded status is NOT True yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(err, "Degraded status is NOT set to True due to wrong CR.")
//save lastTransitionTime of Degraded condition
oldDegradedTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
//save lastTransitionTime of Progressing condition
oldProgressingTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create 2nd CredentialsRequest whose namespace does not exist")
crName2 := "cloud-credential-operator-iam-ro-2"
credentialsRequest2 := credentialsRequest{
name: crName2,
namespace: crNamespace,
provider: providerSpec,
template: crTemp,
}
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName2, "-n", "openshift-cloud-credential-operator", "--ignore-not-found").Execute()
credentialsRequest2.create(oc)
exutil.By("Check 2 CR reporting errors and lastTransitionTime of Degraded and Progressing not changed")
err = wait.Poll(3*time.Second, 60*time.Second, func() (bool, error) {
progressingMessage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].message}`).Output()
if err != nil || !strings.Contains(progressingMessage, "2 reporting errors") {
e2e.Logf("CCO didn't detect 2nd wrong CR yet, and try next round")
return false, nil
}
return true, nil
})
exutil.AssertWaitPollNoErr(err, "CCO didn't detect 2nd wrong CR finally.")
//compare the lastTransitionTime
newDegradedTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Degraded")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(newDegradedTimestamp).To(o.Equal(oldDegradedTimestamp))
newProgressingTimestamp, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "cloud-credential", `-o=jsonpath={.status.conditions[?(@.type=="Progressing")].lastTransitionTime}`).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(newProgressingTimestamp).To(o.Equal(oldProgressingTimestamp))
}) | |||||
test case | openshift/openshift-tests-private | 86102627-1d67-448d-9021-9b071c92421f | NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-High-45219-A fresh cluster should not have stale CR | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-High-45219-A fresh cluster should not have stale CR", func() {
output, _ := oc.AsAdmin().WithoutNamespace().Run("get").Args("service", "controller-manager-service", "-n", "openshift-cloud-credential-operator").Output()
o.Expect(output).To(o.ContainSubstring("Error from server (NotFound)"))
}) | ||||||
test case | openshift/openshift-tests-private | b9fbcaab-7c70-42cc-91fe-099d26868224 | NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-Critical-34470-Cloud credential operator health check | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-ROSA-OSD_CCS-ARO-Author:jshu-Critical-34470-Cloud credential operator health check", func() {
exutil.By("Check CCO status conditions")
//Check CCO mode
mode, err := getCloudCredentialMode(oc)
e2e.Logf("cco mode in cluster is %v", mode)
o.Expect(err).NotTo(o.HaveOccurred())
checkCCOHealth(oc, mode)
exutil.By("Check CCO imagePullPolicy configuration")
imagePullPolicy, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("deployment", "cloud-credential-operator", "-n", "openshift-cloud-credential-operator", "-o=jsonpath={.spec.template.spec.containers[1].imagePullPolicy}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(imagePullPolicy).To(o.Equal("IfNotPresent"))
}) | ||||||
test case | openshift/openshift-tests-private | 8117f02c-198d-4a17-87fd-a84ced9a1b60 | NonHyperShiftHOST-OSD_CCS-ARO-Author:mihuang-LEVEL0-Critical-66538-Azure workload identity cluster healthy check. | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-OSD_CCS-ARO-Author:mihuang-LEVEL0-Critical-66538-Azure workload identity cluster healthy check.", func() {
mode, _ := getCloudCredentialMode(oc)
if !(exutil.CheckPlatform(oc) == "azure" && mode == "manualpodidentity") {
g.Skip("The cluster is not Azure Workload Identity Cluster - skipping test ...")
}
exutil.By("Check CCO status conditions")
checkCCOHealth(oc, mode)
exutil.By("The Azure workload identity cluster does not have `root` credentials")
cmdOut, err := oc.AsAdmin().Run("get").Args("secret", "azure-credentials", "-n", "kube-system").Output()
o.Expect(err).Should(o.HaveOccurred())
o.Expect(cmdOut).To(o.ContainSubstring("Error from server (NotFound)"))
exutil.By("The secret should contain azure_federated_token_file instead of azure credential keys.")
o.Expect(strings.Contains(doOcpReq(oc, "get", true, "secrets", "-n", "openshift-image-registry", "installer-cloud-credentials", "-o=jsonpath={.data}"), "azure_federated_token_file")).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 2b65d5fb-c828-4d1d-a003-3b51ff041b15 | Author:fxie-Critical-64885-CCO-based flow for olm managed operators and AWS STS | ['"path/filepath"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("Author:fxie-Critical-64885-CCO-based flow for olm managed operators and AWS STS", func() {
exutil.SkipIfPlatformTypeNot(oc, "aws")
if !exutil.IsSTSCluster(oc) {
g.Skip("This test case is AWS STS only, skipping")
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "64885"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
stsIAMRoleARN = "whatever"
cloudTokenPath = "anything"
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "AWSProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Inserting an stsIAMRoleARN to the CR")
stsIAMRoleARNPatch := `
spec:
providerSpec:
stsIAMRoleARN: ` + stsIAMRoleARN
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", stsIAMRoleARNPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
stdout, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("extract").
Args("Secret/"+targetSecretName, "-n", targetNs, "--keys", "credentials", "--to", "-").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// The Secret does not contain any sensitive info
e2e.Logf("Secret extracted = %v", stdout)
o.Expect(stdout).To(o.ContainSubstring("[default]"))
o.Expect(stdout).To(o.ContainSubstring("sts_regional_endpoints = regional"))
o.Expect(stdout).To(o.ContainSubstring("role_arn = " + stsIAMRoleARN))
o.Expect(stdout).To(o.ContainSubstring("web_identity_token_file = " + defaultSTSCloudTokenPath))
exutil.By("Inserting a cloudTokenPath to the CR")
cloudTokenPathPatch := `
spec:
cloudTokenPath: ` + cloudTokenPath
err = oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", cloudTokenPathPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is updated in the correct way")
o.Eventually(func() bool {
stdout, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("extract").
Args("Secret/"+targetSecretName, "-n", targetNs, "--keys", "credentials", "--to", "-").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// The Secret does not contain any sensitive info
e2e.Logf("Secret extracted = %v", stdout)
return strings.Contains(stdout, "web_identity_token_file = "+cloudTokenPath)
}).WithTimeout(DefaultTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 4eea3a0c-2288-48c2-805d-ba6987f12f57 | NonHyperShiftHOST-OSD_CCS-ARO-Author:jshu-Critical-69971-Azure workload identity management for olm managed operators | ['"path/filepath"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-OSD_CCS-ARO-Author:jshu-Critical-69971-Azure workload identity management for olm managed operators", func() {
exutil.SkipIfPlatformTypeNot(oc, "azure")
if !exutil.IsWorkloadIdentityCluster(oc) {
g.Skip("This test case is for Azure Workload Identity only, skipping")
}
//Provide the following Azure Credentials with fake values
azureCredList := []azureCredential{
{
key: "azure_subscription_id",
value: "12345678-1234-1234-1234-123456789ab",
},
{
key: "azure_tenant_id",
value: "23456789-2345-2345-2345-23456789abcd",
},
{
key: "azure_region",
value: "eastus",
},
{
key: "azure_client_id",
value: "3456789a-3456-3456-3456-23456789abcde",
},
{
key: "azure_federated_token_file",
value: "/var/run/secrets/token",
},
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "69971"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating the dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "AzureProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Patching the Azure Credentials and cloudTokenPath to the CR")
crPatch := `
spec:
cloudTokenPath: ` + azureCredList[4].value + `
providerSpec:
azureSubscriptionID: ` + azureCredList[0].value + `
azureTenantID: ` + azureCredList[1].value + `
azureRegion: ` + azureCredList[2].value + `
azureClientID: ` + azureCredList[3].value
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", crPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
for _, azureCred := range azureCredList {
credential, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/"+targetSecretName, "-n", targetNs, "--keys", azureCred.key, "--to", "-").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(credential).To(o.ContainSubstring(azureCred.value))
}
}) | |||||
test case | openshift/openshift-tests-private | ca39ac8e-dc06-44a3-b04d-9c27fc4ad4a5 | Author:jshu-NonHyperShiftHOST-OSD_CCS-Critical-75429-GCP workload identity management for olm managed operators | ['"encoding/base64"', '"path/filepath"', '"strings"', '"time"', '"github.com/tidwall/gjson"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("Author:jshu-NonHyperShiftHOST-OSD_CCS-Critical-75429-GCP workload identity management for olm managed operators", func() {
exutil.SkipIfPlatformTypeNot(oc, "gcp")
if !exutil.IsWorkloadIdentityCluster(oc) {
g.Skip("This test case is for GCP Workload Identity only, skipping")
}
//Provide the following GCP Credentials with fake values
gcpCredList := []gcpCredential{
{
key: "audience",
value: "//iam.googleapis.com/projects/1042363005003/locations/global/workloadIdentityPools/cco-test/providers/cco-test",
},
{
key: "serviceAccountEmail",
value: "[email protected]",
},
{
key: "cloudTokenPath",
value: "/var/run/secrets/token",
},
}
var (
testDataDir = exutil.FixturePath("testdata", "cluster_operator/cloudcredential")
testCaseID = "75429"
crName = "cr-" + testCaseID
targetSecretName = crName
targetNs = oc.Namespace()
)
var (
targetSecretCreated = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", "-n", targetNs).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Contains(stdout, targetSecretName)
}
)
exutil.By("Creating the dummy CR")
cr := credentialsRequest{
name: crName,
namespace: targetNs,
provider: "GCPProviderSpec",
template: filepath.Join(testDataDir, "credentials_request.yaml"),
}
defer func() {
_ = oc.AsAdmin().WithoutNamespace().Run("delete").Args("CredentialsRequest", crName, "-n", ccoNs).Execute()
}()
cr.create(oc)
exutil.By("Making sure the target Secret is not created")
o.Consistently(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeFalse())
exutil.By("Patching the GCP Credentials and cloudTokenPath to the CR")
crPatch := `
spec:
cloudTokenPath: ` + gcpCredList[2].value + `
providerSpec:
audience: ` + gcpCredList[0].value + `
serviceAccountEmail: ` + gcpCredList[1].value
err := oc.
AsAdmin().
WithoutNamespace().
Run("patch").
Args("CredentialsRequest", crName, "-n", ccoNs, "--type", "merge", "-p", crPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Making sure the target Secret is created correctly")
o.Eventually(targetSecretCreated).WithTimeout(60 * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
credentialBase64, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", targetSecretName, "-n", targetNs, "-o=jsonpath={.data.service_account\\.json}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
credential, err := base64.StdEncoding.DecodeString(credentialBase64)
o.Expect(err).NotTo(o.HaveOccurred())
//compare audience
gen_audience := gjson.Get(string(credential), `audience`).String()
o.Expect(gen_audience).To(o.Equal(gcpCredList[0].value))
//check serviceAccountEmail
gen_service_account := gjson.Get(string(credential), `service_account_impersonation_url`).String()
o.Expect(gen_service_account).To(o.ContainSubstring(gcpCredList[1].value))
//compare token path
gen_token_path := gjson.Get(string(credential), `credential_source.file`).String()
o.Expect(gen_token_path).To(o.Equal(gcpCredList[2].value))
}) | |||||
test case | openshift/openshift-tests-private | 862f336e-91c3-4714-8643-dff4a2bbea85 | NonHyperShiftHOST-Author:fxie-Critical-68220-Leverage Composable OpenShift feature to make cloud-credential optional | ['"context"', '"fmt"', '"io"', '"net/http"', '"os"', '"os/exec"', '"strings"', '"github.com/google/go-github/v57/github"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/cloudcredential/cloudcredential.go | g.It("NonHyperShiftHOST-Author:fxie-Critical-68220-Leverage Composable OpenShift feature to make cloud-credential optional", func() {
var (
getManifestContent = func(manifest *github.RepositoryContent) []byte {
// Prefer manifest.Content
if content, _ := manifest.GetContent(); content != "" {
return []byte(content)
}
// Fall back to downloadURL
downloadURL := manifest.GetDownloadURL()
o.Expect(downloadURL).NotTo(o.BeEmpty())
req, err := http.NewRequest("GET", downloadURL, nil)
o.Expect(err).NotTo(o.HaveOccurred())
res, err := http.DefaultClient.Do(req)
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
_ = res.Body.Close()
}()
content, err := io.ReadAll(res.Body)
o.Expect(err).NotTo(o.HaveOccurred())
return content
}
/*
Here we avoid deserializing manifests through a runtime.Scheme since
installing all required types (to the scheme) would bring in quite a few
dependencies, making our test repo unnecessarily heavy.
*/
resourceInfoFromManifest = func(manifest []byte) (string, string, string) {
var deserializedManifest map[string]any
err := yaml.Unmarshal(manifest, &deserializedManifest)
o.Expect(err).NotTo(o.HaveOccurred())
groupVersion, ok := deserializedManifest["apiVersion"].(string)
o.Expect(ok).To(o.BeTrue())
groupVersionSlice := strings.Split(groupVersion, "/")
kind, ok := deserializedManifest["kind"].(string)
o.Expect(ok).To(o.BeTrue())
// The oc client is smart enough to map kind to resource before making an API call.
// There's no need to query the discovery endpoint of the API server ourselves to obtain the gvr.
var resourceType, namespace, name string
switch len(groupVersionSlice) {
// The resource is a part of the core group
case 1:
resourceType = kind
// The resource is not a part of the core group
case 2:
resourceType = fmt.Sprintf("%s.%s.%s", kind, groupVersionSlice[1], groupVersionSlice[0])
default:
e2e.Failf("Unexpected apiVersion format")
}
metadata, ok := deserializedManifest["metadata"].(map[string]any)
o.Expect(ok).To(o.BeTrue())
if _, isNamespaced := metadata["namespace"]; isNamespaced {
namespace, ok = metadata["namespace"].(string)
o.Expect(ok).To(o.BeTrue())
}
name, ok = metadata["name"].(string)
o.Expect(ok).To(o.BeTrue())
e2e.Logf("Resource type = %v, namespace = %v, name = %v", resourceType, namespace, name)
return resourceType, namespace, name
}
)
// Get GitHub client
ghClient := github.NewClient(nil)
// Authenticate for a much larger rate limit
if ghToken := os.Getenv("GITHUB_TOKEN"); ghToken != "" {
ghClient = ghClient.WithAuthToken(ghToken)
}
// Get cluster version
majorMinorVersion, _, err := exutil.GetClusterVersion(oc)
o.Expect(err).NotTo(o.HaveOccurred())
targetBranch := fmt.Sprintf("release-%s", majorMinorVersion)
// Get manifest dir of the CCO repo
// There's no need to verify the absence of CredentialsRequests defined in other repos.
// We only need to make sure the corresponding CRD is not found in-cluster.
e2e.Logf("Listing manifest directory of branch %v", targetBranch)
_, dir, _, err := ghClient.
Repositories.
GetContents(context.Background(), "openshift", ccoRepo, ccoManifestPath, &github.RepositoryContentGetOptions{
Ref: targetBranch,
})
o.Expect(err).NotTo(o.HaveOccurred())
// Iterate through items in the manifest dir
for _, manifest := range dir {
if manifest.GetType() != "file" {
continue
}
manifestName := manifest.GetName()
if !strings.HasSuffix(manifestName, ".yaml") && !strings.HasSuffix(manifestName, ".yml") {
continue
}
e2e.Logf("Getting content of manifest %v", manifestName)
content := getManifestContent(manifest)
e2e.Logf("Extracting resource info from manifest")
resourceType, namespace, name := resourceInfoFromManifest(content)
e2e.Logf("Requesting manifest against the API server")
getReqArgs := []string{resourceType, name}
if namespace != "" {
getReqArgs = append(getReqArgs, "-n", namespace)
}
// err is the error returned by executing an exec.Command
// stderr captures the original error message return by the API server
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("get").Args(getReqArgs...).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.Or(
o.ContainSubstring("not found"),
o.ContainSubstring("the server doesn't have a resource type"),
))
}
}) | |||||
test | openshift/openshift-tests-private | e3c17a06-8a37-445a-848e-4dc612bcd8b0 | hive | import (
"encoding/json"
"path/filepath"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
e2e "k8s.io/kubernetes/test/e2e/framework"
) | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive.go | package hive
import (
"encoding/json"
"path/filepath"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
e2e "k8s.io/kubernetes/test/e2e/framework"
)
//
// Hive test case suite for platform independent and all other platforms
//
var _ = g.Describe("[sig-hive] Cluster_Operator hive should", func() {
defer g.GinkgoRecover()
var (
oc = exutil.NewCLI("hive", exutil.KubeConfigPath())
ns hiveNameSpace
og operatorGroup
sub subscription
hc hiveconfig
testDataDir string
iaasPlatform string
testOCPImage string
)
g.BeforeEach(func() {
// skip ARM64 arch
architecture.SkipNonAmd64SingleArch(oc)
//Install Hive operator if not
testDataDir = exutil.FixturePath("testdata", "cluster_operator/hive")
_, _ = installHiveOperator(oc, &ns, &og, &sub, &hc, testDataDir)
// get IaaS platform
iaasPlatform = exutil.CheckPlatform(oc)
//Get OCP Image for Hive testing
testOCPImage = getTestOCPImage()
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "42345"|./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-42345-Low-42349-shouldn't create provisioning pod if region mismatch in install config vs Cluster Deployment [Serial]", func() {
testCaseID := "42345"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
switch iaasPlatform {
case "aws":
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion2,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
case "gcp":
exutil.By("Config GCP Install-Config Secret...")
projectID, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure/cluster", "-o=jsonpath={.status.platformStatus.gcp.projectID}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(projectID).NotTo(o.BeEmpty())
installConfigSecret := gcpInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: GCPBaseDomain,
name2: cdName,
region: GCPRegion2,
projectid: projectID,
template: filepath.Join(testDataDir, "gcp-install-config.yaml"),
}
exutil.By("Config GCP ClusterDeployment...")
cluster := gcpClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: GCPBaseDomain,
clusterName: cdName,
platformType: "gcp",
credRef: GCPCreds,
region: GCPRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment-gcp.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
case "azure":
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion2,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
default:
g.Skip("unsupported ClusterDeployment type")
}
exutil.By("Check provision pod can't be created")
watchProvisionpod := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pods", "-n", oc.Namespace()).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(stdout, "-provision-") {
e2e.Logf("Provision pod should not be created")
return false
}
return true
}
o.Consistently(watchProvisionpod).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check conditions of ClusterDeployment, the type RequirementsMet should be False")
waitForClusterDeploymentRequirementsMetFail := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "RequirementsMet")
if status, ok := condition["status"]; !ok || status != "False" {
e2e.Logf("For condition RequirementsMet, expected status is False, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "InstallConfigValidationFailed" {
e2e.Logf("For condition RequirementsMet, expected reason is InstallConfigValidationFailed, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || message != "install config region does not match cluster deployment region" {
e2e.Logf("For condition RequirementsMet, expected message is \ninstall config region does not match cluster deployment region, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition RequirementsMet, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClusterDeploymentRequirementsMetFail).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-42349: Sort clusterdeployment conditions")
var ClusterDeploymentConditions []map[string]string
checkConditionSequence := func() bool {
stdout, _, err := oc.AsAdmin().Run("get").Args("ClusterDeployment", cdName, "-o", "jsonpath={.status.conditions}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &ClusterDeploymentConditions)
o.Expect(err).NotTo(o.HaveOccurred())
if conditionType0, ok := ClusterDeploymentConditions[0]["type"]; !ok || conditionType0 != "RequirementsMet" {
e2e.Logf("Error! condition RequirementsMet is not at the top of the conditions list")
return false
}
if conditionStatus0, ok := ClusterDeploymentConditions[0]["status"]; !ok || conditionStatus0 != "False" {
e2e.Logf("Error! condition RequirementsMet is not in False status")
return false
}
e2e.Logf("Check if conditions with desired state are at the middle, conditions with Unknown are at the bottom")
conditionNum := len(ClusterDeploymentConditions)
findUnknownFlag := false
for i := 1; i < conditionNum; i++ {
conditionStatus, ok := ClusterDeploymentConditions[i]["status"]
if !ok {
e2e.Logf("Error! a condition doesn't have status")
return false
}
if conditionStatus == "Unknown" {
findUnknownFlag = true
} else {
if findUnknownFlag {
e2e.Logf("condition with Unknown is not at the bottom")
return false
}
}
}
e2e.Logf("Check is passed! All conditions with desired state are at the middle, and all conditions with Unknown are at the bottom")
return true
}
o.Consistently(checkConditionSequence).WithTimeout(2 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Critical-29670-install/uninstall hive operator from OperatorHub", func() {
exutil.By("Check Subscription...")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "AllCatalogSourcesHealthy", ok, DefaultTimeout, []string{"sub", sub.name, "-n",
sub.namespace, "-o=jsonpath={.status.conditions[0].reason}"}).check(oc)
exutil.By("Check Hive Operator pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check Hive Operator pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Hive Operator sucessfully installed !!! ")
exutil.By("Check hive-clustersync pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hive-clustersync pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Check hive-controllers pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hive-controllers pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Check hiveadmission pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hiveadmission pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission", "-n",
sub.namespace, "-o=jsonpath={.items[*].status.phase}"}).check(oc)
exutil.By("Hive controllers,clustersync and hiveadmission sucessfully installed !!! ")
})
//author: [email protected]
//default duration is 15m for extended-platform-tests and 35m for jenkins job, need to reset for ClusterPool and ClusterDeployment cases
//example: ./bin/extended-platform-tests run all --dry-run|grep "41932"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Medium-41932-Add metric for hive-operator[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
exutil.By("Check hive-operator metrics can be queried from thanos-querier")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query1 := "hive_operator_reconcile_seconds_sum"
query2 := "hive_operator_reconcile_seconds_count"
query3 := "hive_operator_reconcile_seconds_bucket"
query4 := "hive_hiveconfig_conditions"
query := []string{query1, query2, query3, query4}
checkMetricExist(oc, ok, token, thanosQuerierURL, query)
exutil.By("Check HiveConfig status from Metric...")
expectedType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("HiveConfig", "hive", "-o=jsonpath={.status.conditions[0].type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
expectedReason, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("HiveConfig", "hive", "-o=jsonpath={.status.conditions[0].reason}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
checkHiveConfigMetric(oc, "condition", expectedType, token, thanosQuerierURL, query4)
checkHiveConfigMetric(oc, "reason", expectedReason, token, thanosQuerierURL, query4)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "55904"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Low-55904-Hiveadmission log enhancement[Serial]", func() {
hiveadmissionPod := getHiveadmissionPod(oc, sub.namespace)
hiveadmissionPodLog, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveadmissionPod, "-n", sub.namespace).Output()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(hiveadmissionPodLog, "failed to list") {
e2e.Failf("the pod log includes failed to list")
}
if !strings.Contains(hiveadmissionPodLog, "Running API Priority and Fairness config worker") {
e2e.Failf("the pod log does not include Running API Priority and Fairness config worker")
}
})
})
| package hive | ||||
test case | openshift/openshift-tests-private | 2e21602f-fc98-4482-86e9-c6025490dbfc | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-42345-Low-42349-shouldn't create provisioning pod if region mismatch in install config vs Cluster Deployment [Serial] | ['"encoding/json"', '"path/filepath"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-42345-Low-42349-shouldn't create provisioning pod if region mismatch in install config vs Cluster Deployment [Serial]", func() {
testCaseID := "42345"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
switch iaasPlatform {
case "aws":
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion2,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
case "gcp":
exutil.By("Config GCP Install-Config Secret...")
projectID, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure/cluster", "-o=jsonpath={.status.platformStatus.gcp.projectID}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(projectID).NotTo(o.BeEmpty())
installConfigSecret := gcpInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: GCPBaseDomain,
name2: cdName,
region: GCPRegion2,
projectid: projectID,
template: filepath.Join(testDataDir, "gcp-install-config.yaml"),
}
exutil.By("Config GCP ClusterDeployment...")
cluster := gcpClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: GCPBaseDomain,
clusterName: cdName,
platformType: "gcp",
credRef: GCPCreds,
region: GCPRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment-gcp.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
case "azure":
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion2,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
default:
g.Skip("unsupported ClusterDeployment type")
}
exutil.By("Check provision pod can't be created")
watchProvisionpod := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("pods", "-n", oc.Namespace()).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(stdout, "-provision-") {
e2e.Logf("Provision pod should not be created")
return false
}
return true
}
o.Consistently(watchProvisionpod).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check conditions of ClusterDeployment, the type RequirementsMet should be False")
waitForClusterDeploymentRequirementsMetFail := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "RequirementsMet")
if status, ok := condition["status"]; !ok || status != "False" {
e2e.Logf("For condition RequirementsMet, expected status is False, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "InstallConfigValidationFailed" {
e2e.Logf("For condition RequirementsMet, expected reason is InstallConfigValidationFailed, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || message != "install config region does not match cluster deployment region" {
e2e.Logf("For condition RequirementsMet, expected message is \ninstall config region does not match cluster deployment region, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition RequirementsMet, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClusterDeploymentRequirementsMetFail).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-42349: Sort clusterdeployment conditions")
var ClusterDeploymentConditions []map[string]string
checkConditionSequence := func() bool {
stdout, _, err := oc.AsAdmin().Run("get").Args("ClusterDeployment", cdName, "-o", "jsonpath={.status.conditions}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &ClusterDeploymentConditions)
o.Expect(err).NotTo(o.HaveOccurred())
if conditionType0, ok := ClusterDeploymentConditions[0]["type"]; !ok || conditionType0 != "RequirementsMet" {
e2e.Logf("Error! condition RequirementsMet is not at the top of the conditions list")
return false
}
if conditionStatus0, ok := ClusterDeploymentConditions[0]["status"]; !ok || conditionStatus0 != "False" {
e2e.Logf("Error! condition RequirementsMet is not in False status")
return false
}
e2e.Logf("Check if conditions with desired state are at the middle, conditions with Unknown are at the bottom")
conditionNum := len(ClusterDeploymentConditions)
findUnknownFlag := false
for i := 1; i < conditionNum; i++ {
conditionStatus, ok := ClusterDeploymentConditions[i]["status"]
if !ok {
e2e.Logf("Error! a condition doesn't have status")
return false
}
if conditionStatus == "Unknown" {
findUnknownFlag = true
} else {
if findUnknownFlag {
e2e.Logf("condition with Unknown is not at the bottom")
return false
}
}
}
e2e.Logf("Check is passed! All conditions with desired state are at the middle, and all conditions with Unknown are at the bottom")
return true
}
o.Consistently(checkConditionSequence).WithTimeout(2 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 549bfd4e-2c0c-4c61-a123-c7532cf45e06 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Critical-29670-install/uninstall hive operator from OperatorHub | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Critical-29670-install/uninstall hive operator from OperatorHub", func() {
exutil.By("Check Subscription...")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "AllCatalogSourcesHealthy", ok, DefaultTimeout, []string{"sub", sub.name, "-n",
sub.namespace, "-o=jsonpath={.status.conditions[0].reason}"}).check(oc)
exutil.By("Check Hive Operator pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check Hive Operator pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Hive Operator sucessfully installed !!! ")
exutil.By("Check hive-clustersync pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hive-clustersync pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Check hive-controllers pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hive-controllers pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager", "-n",
sub.namespace, "-o=jsonpath={.items[0].status.phase}"}).check(oc)
exutil.By("Check hiveadmission pods are created !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
exutil.By("Check hiveadmission pods are in running state !!!")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission", "-n",
sub.namespace, "-o=jsonpath={.items[*].status.phase}"}).check(oc)
exutil.By("Hive controllers,clustersync and hiveadmission sucessfully installed !!! ")
}) | ||||||
test case | openshift/openshift-tests-private | 0bd87227-64dc-4d74-85df-9a7d371841d9 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Medium-41932-Add metric for hive-operator[Serial] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-Medium-41932-Add metric for hive-operator[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
exutil.By("Check hive-operator metrics can be queried from thanos-querier")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query1 := "hive_operator_reconcile_seconds_sum"
query2 := "hive_operator_reconcile_seconds_count"
query3 := "hive_operator_reconcile_seconds_bucket"
query4 := "hive_hiveconfig_conditions"
query := []string{query1, query2, query3, query4}
checkMetricExist(oc, ok, token, thanosQuerierURL, query)
exutil.By("Check HiveConfig status from Metric...")
expectedType, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("HiveConfig", "hive", "-o=jsonpath={.status.conditions[0].type}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
expectedReason, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("HiveConfig", "hive", "-o=jsonpath={.status.conditions[0].reason}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
checkHiveConfigMetric(oc, "condition", expectedType, token, thanosQuerierURL, query4)
checkHiveConfigMetric(oc, "reason", expectedReason, token, thanosQuerierURL, query4)
}) | ||||||
test case | openshift/openshift-tests-private | 6d74d47b-7b63-470d-a7b7-601ebb4ddbab | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Low-55904-Hiveadmission log enhancement[Serial] | ['"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Low-55904-Hiveadmission log enhancement[Serial]", func() {
hiveadmissionPod := getHiveadmissionPod(oc, sub.namespace)
hiveadmissionPodLog, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveadmissionPod, "-n", sub.namespace).Output()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(hiveadmissionPodLog, "failed to list") {
e2e.Failf("the pod log includes failed to list")
}
if !strings.Contains(hiveadmissionPodLog, "Running API Priority and Fairness config worker") {
e2e.Failf("the pod log does not include Running API Priority and Fairness config worker")
}
}) | |||||
test | openshift/openshift-tests-private | b7e0ed1c-88ca-48f3-bba8-270aa72a085f | hive_azure | import (
"fmt"
"io/ioutil"
"os"
"path/filepath"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
e2e "k8s.io/kubernetes/test/e2e/framework"
) | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | package hive
import (
"fmt"
"io/ioutil"
"os"
"path/filepath"
"strings"
"time"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
e2e "k8s.io/kubernetes/test/e2e/framework"
)
//
// Hive test case suite for Azure
//
var _ = g.Describe("[sig-hive] Cluster_Operator hive should", func() {
defer g.GinkgoRecover()
var (
oc = exutil.NewCLI("hive", exutil.KubeConfigPath())
ns hiveNameSpace
og operatorGroup
sub subscription
hc hiveconfig
testDataDir string
testOCPImage string
region string
basedomain string
cloudName string
isGovCloud bool
)
g.BeforeEach(func() {
// Skip ARM64 arch
architecture.SkipNonAmd64SingleArch(oc)
// Skip if running on a non-Azure platform
exutil.SkipIfPlatformTypeNot(oc, "azure")
// Install Hive operator if non-existent
testDataDir = exutil.FixturePath("testdata", "cluster_operator/hive")
_, _ = installHiveOperator(oc, &ns, &og, &sub, &hc, testDataDir)
// Get OCP Image for Hive testing
testOCPImage = getTestOCPImage()
// Get platform configurations
region = getRegion(oc)
basedomain = getBasedomain(oc)
isGovCloud = strings.Contains(region, "usgov")
cloudName = AzurePublic
if isGovCloud {
e2e.Logf("Running on MAG")
cloudName = AzureGov
}
})
// Author: [email protected] [email protected]
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-High-25447-High-28657-High-45175-[Mag] Hive API support for Azure [Serial]", func() {
testCaseID := "25447"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-worker-azure.yaml")
inframachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-infra-azure.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAzureTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAzureTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check Azure ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-28657: Hive supports remote Machine Set Management for Azure")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check worker machinepool .status.replicas = 3")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check only 1 machineset up")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check only one machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 3")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 3}}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
machinesetsArray := strings.Fields(machinesetsname)
o.Expect(len(machinesetsArray) == 3).Should(o.BeTrue())
for _, machinesetName := range machinesetsArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinesetName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
}
e2e.Logf("Check machinesets scale up to 3")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 3 machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 2")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 2}}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
machinesetsArray = strings.Fields(machinesetsname)
o.Expect(len(machinesetsArray) == 2).Should(o.BeTrue())
for _, machinesetName := range machinesetsArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinesetName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
}
e2e.Logf("Check machinesets scale down to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 2 machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
})
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-33854-Hive supports Azure ClusterPool [Serial]", func() {
testCaseID := "33854"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool-azure.yaml")
pool := azureClusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AzureBaseDomain,
imageSetRef: imageSetName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if Azure ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, AzureClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Check if CD is Hibernating")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
exutil.By("Patch pool.spec.lables.test=test...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"labels":{"test":"test"}}}`}).check(oc)
exutil.By("The existing CD in the pool has no test label")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "test", nok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.labels}"}).check(oc)
}
exutil.By("The new CD in the pool should have the test label")
e2e.Logf("Delete the old CD in the pool")
newCheck("expect", "delete", asAdmin, withoutNamespace, contain, "delete", ok, ClusterUninstallTimeout, []string{"ClusterDeployment", cdArray[0], "-n", cdArray[0]}).check(oc)
e2e.Logf("Get the CD list from the pool again.")
cdListStr = getCDlistfromPool(oc, poolName)
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "test", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.labels}"}).check(oc)
}
})
//author: [email protected]
//default duration is 15m for extended-platform-tests and 35m for jenkins job, need to reset for ClusterPool and ClusterDeployment cases
//example: ./bin/extended-platform-tests run all --dry-run|grep "35297"|./bin/extended-platform-tests run --timeout 90m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-35297-Hive supports cluster hibernation[Serial]", func() {
testCaseID := "35297"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Azure ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check CD has Hibernating condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
exutil.By("patch the CD to Hibernating...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Hibernating"}}`}).check(oc)
e2e.Logf("Wait for CD to be Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
exutil.By("patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "44946"|./bin/extended-platform-tests run --timeout 90m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-44946-Keep it hot when HibernateAfter is setting [Serial]", func() {
testCaseID := "44946"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Step1 : Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool-azure.yaml")
pool := azureClusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AzureBaseDomain,
imageSetRef: imageSetName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 1,
maxConcurrent: 1,
hibernateAfter: "5m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Step2 : Check if Azure ClusterPool created successfully and become ready")
//runningCount is 1 so pool status should be ready: 1
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, AzureClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.ready}"}).check(oc)
exutil.By("Get cd name from cdlist")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
o.Expect(len(cdArray)).Should(o.BeNumerically(">=", 1))
oldCdName := cdArray[0]
exutil.By("Step4 : wait 7 mins")
//hibernateAfter is 5 min, wait for > 5 min (Need wait hardcode timer instead of check any condition here so use Sleep function directly) to check the cluster is still running status.
time.Sleep((HibernateAfterTimer + DefaultTimeout) * time.Second)
exutil.By("Step6 and Step7 : Check unclaimed cluster is still in Running status after hibernateAfter time, because of runningcount=1")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterDeployment", oldCdName, "-n", oldCdName}).check(oc)
exutil.By("Step8 :Check installedTimestap and hibernating lastTransitionTime filed.")
installedTimestap, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.installedTimestamp}`))
o.Expect(err).NotTo(o.HaveOccurred())
hibernateTransition, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].lastTransitionTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
difference := hibernateTransition.Sub(installedTimestap)
e2e.Logf("Timestamp difference is %v min", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically("<=", 1))
exutil.By("Step4 : Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
claimedTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterClaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.creationTimestamp}"))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get clusterclaim timestamp,creationTimestamp is %s", claimedTimestamp)
exutil.By("Check if ClusterClaim created successfully and become running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace(), claimName}).check(oc)
exutil.By("Check if CD is Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, 2*ClusterResumeTimeout, []string{"ClusterDeployment", oldCdName, "-n", oldCdName}).check(oc)
hibernateTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].lastProbeTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get hibernattimestaps, hibernateTimestamp is %s", hibernateTimestamp)
exutil.By("Step5 : Get Timestamp difference")
difference = hibernateTimestamp.Sub(claimedTimestamp)
e2e.Logf("Timestamp difference is %v mins", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically(">=", 5))
})
//author: [email protected]
//default duration is 15m for extended-platform-tests and 35m for jenkins job, need to reset for ClusterPool and ClusterDeployment cases
//example: ./bin/extended-platform-tests run all --dry-run|grep "52415"|./bin/extended-platform-tests run --timeout 90m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-52415-[Azure]Hive Machinepool test for autoscale [Serial]", func() {
testCaseID := "52415"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create infra MachinePool ...")
inframachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-infra-azure.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAzureTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Patch static replicas to autoscaler")
exutil.By("OCP-52415: [Azure]Allow minReplicas autoscaling of MachinePools to be 0")
e2e.Logf("Check hive allow set minReplicas=0 without zone setting")
autoScalingMax := "3"
autoScalingMin := "0"
removeConfig := "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig := fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 2*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check hive allow set minReplicas=0 within zone setting")
cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
infra2MachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-infra2
namespace: ` + oc.Namespace() + `
spec:
autoscaling:
maxReplicas: 3
minReplicas: 0
clusterDeploymentRef:
name: ` + cdName + `
labels:
node-role.kubernetes.io: infra2
node-role.kubernetes.io/infra2: ""
name: infra2
platform:
azure:
osDisk:
diskSizeGB: 128
type: Standard_D4s_v3
zones:
- "1"
- "2"
- "3"`
var filename = testCaseID + "-machinepool-infra2.yaml"
err = ioutil.WriteFile(filename, []byte(infra2MachinepoolYaml), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("-f", filename, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 2*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
exutil.By("Check Hive supports autoscale for Azure")
patchYaml := `
spec:
scaleDown:
enabled: true
delayAfterAdd: 10s
delayAfterDelete: 10s
delayAfterFailure: 10s
unneededTime: 10s`
e2e.Logf("Add busybox in remote cluster and check machines will scale up to maxReplicas")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ClusterAutoscaler", "default", "--type", "merge", "-p", patchYaml}).check(oc)
workloadYaml := filepath.Join(testDataDir, "workload.yaml")
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "busybox", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Deployment", "busybox", "-n", "default"}).check(oc)
e2e.Logf("Check replicas will scale up to maximum value")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Delete busybox in remote cluster and check machines will scale down to minReplicas")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas will scale down to minimum value")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "54048"|./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Medium-54048-[Mag] Hive to support cli-domain-from-installer-image annotation [Serial]", func() {
testCaseID := "54048"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "azure-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := azureInstallConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Get domain from installerimage...")
clusterVersion, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterversion/version", "-o=jsonpath={.status.desired.version}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(clusterVersion).NotTo(o.BeEmpty())
originalInstallerImage, err := getPullSpec(oc, "installer", clusterVersion)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(originalInstallerImage).NotTo(o.BeEmpty())
e2e.Logf("ClusterVersion is %s, originalInstallerImage is %s", clusterVersion, originalInstallerImage)
installerDomain := strings.SplitN(originalInstallerImage, "/", 2)[0]
installerPath := strings.SplitN(originalInstallerImage, "/", 2)[1]
overrideDomain := "mihuang.io"
overrideInstallerImage := overrideDomain + "/" + installerPath
type imageOverride struct {
copyCliDomain string
installerImageOverride string
}
var scenarios = []imageOverride{
{
"true",
overrideInstallerImage,
},
{
"false",
overrideInstallerImage,
},
{
"true",
"",
},
}
for i := 0; i < len(scenarios); i++ {
func() {
//create cluster
if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=true and installerImageOverride set...")
} else if scenarios[i].copyCliDomain == "false" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=false and installerImageOverride set...")
} else if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == "" {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=true and installerImageOverride unset...")
}
cluster := azureClusterDeployment{
fake: "false",
copyCliDomain: scenarios[i].copyCliDomain,
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
installerImageOverride: scenarios[i].installerImageOverride,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
//get conditions
exutil.By("Check if provisioning pod create success")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Provisioning", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check cd .status.installerImage")
installerImage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installerImage}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
overrideImageDomain := strings.SplitN(installerImage, "/", 2)[0]
e2e.Logf("overrideImageDomain is %s", overrideImageDomain)
e2e.Logf("Check cd .status.cliImage")
cliImage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.cliImage}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
cliImageDomain := strings.SplitN(cliImage, "/", 2)[0]
e2e.Logf("cliImageDomain is %s", cliImageDomain)
//check conditions for scenarios
if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Check if both cliImage and installerImage use the new domain")
o.Expect(overrideImageDomain).To(o.Equal(overrideDomain))
o.Expect(cliImageDomain).To(o.Equal(overrideDomain))
} else if scenarios[i].copyCliDomain == "false" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Check if cliImage use offical domain and installerImage use the new domain")
o.Expect(overrideImageDomain).To(o.Equal(overrideDomain))
o.Expect(cliImageDomain).To(o.Equal(installerDomain))
} else {
exutil.By("Check if both cliImage and installerImage use the offical image")
o.Expect(overrideImageDomain).To(o.Equal(installerDomain))
o.Expect(cliImageDomain).To(o.Equal(installerDomain))
}
}()
}
})
})
| package hive | ||||
test case | openshift/openshift-tests-private | d61bb1ac-39e6-450d-af68-760f4127cb96 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-High-25447-High-28657-High-45175-[Mag] Hive API support for Azure [Serial] | ['"os"', '"path/filepath"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-High-25447-High-28657-High-45175-[Mag] Hive API support for Azure [Serial]", func() {
testCaseID := "25447"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-worker-azure.yaml")
inframachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-infra-azure.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAzureTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAzureTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check Azure ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-28657: Hive supports remote Machine Set Management for Azure")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check worker machinepool .status.replicas = 3")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check only 1 machineset up")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check only one machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 3")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 3}}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
machinesetsArray := strings.Fields(machinesetsname)
o.Expect(len(machinesetsArray) == 3).Should(o.BeTrue())
for _, machinesetName := range machinesetsArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinesetName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
}
e2e.Logf("Check machinesets scale up to 3")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 3 machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 2")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 2}}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
machinesetsArray = strings.Fields(machinesetsname)
o.Expect(len(machinesetsArray) == 2).Should(o.BeTrue())
for _, machinesetName := range machinesetsArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinesetName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
}
e2e.Logf("Check machinesets scale down to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 2 machines in Running status")
// Can't filter by infra label because of bug https://issues.redhat.com/browse/HIVE-1922
//newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[?(@.spec.metadata.labels.node-role\\.kubernetes\\.io==\"infra\")].status.phase}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 7deddea2-6bb9-45a9-9684-b6aa21d6874a | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-33854-Hive supports Azure ClusterPool [Serial] | ['"path/filepath"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-33854-Hive supports Azure ClusterPool [Serial]", func() {
testCaseID := "33854"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool-azure.yaml")
pool := azureClusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AzureBaseDomain,
imageSetRef: imageSetName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if Azure ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, AzureClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Check if CD is Hibernating")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
exutil.By("Patch pool.spec.lables.test=test...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"labels":{"test":"test"}}}`}).check(oc)
exutil.By("The existing CD in the pool has no test label")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "test", nok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.labels}"}).check(oc)
}
exutil.By("The new CD in the pool should have the test label")
e2e.Logf("Delete the old CD in the pool")
newCheck("expect", "delete", asAdmin, withoutNamespace, contain, "delete", ok, ClusterUninstallTimeout, []string{"ClusterDeployment", cdArray[0], "-n", cdArray[0]}).check(oc)
e2e.Logf("Get the CD list from the pool again.")
cdListStr = getCDlistfromPool(oc, poolName)
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "test", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.labels}"}).check(oc)
}
}) | |||||
test case | openshift/openshift-tests-private | 124663db-e30b-4829-bc9a-abe7a759f58e | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-35297-Hive supports cluster hibernation[Serial] | ['"path/filepath"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-35297-Hive supports cluster hibernation[Serial]", func() {
testCaseID := "35297"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Azure ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check CD has Hibernating condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
exutil.By("patch the CD to Hibernating...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Hibernating"}}`}).check(oc)
e2e.Logf("Wait for CD to be Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
exutil.By("patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 18d03190-37f7-4bbd-b6b4-f020264d8ae0 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-44946-Keep it hot when HibernateAfter is setting [Serial] | ['"path/filepath"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-44946-Keep it hot when HibernateAfter is setting [Serial]", func() {
testCaseID := "44946"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Step1 : Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool-azure.yaml")
pool := azureClusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AzureBaseDomain,
imageSetRef: imageSetName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 1,
maxConcurrent: 1,
hibernateAfter: "5m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Step2 : Check if Azure ClusterPool created successfully and become ready")
//runningCount is 1 so pool status should be ready: 1
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, AzureClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.ready}"}).check(oc)
exutil.By("Get cd name from cdlist")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
o.Expect(len(cdArray)).Should(o.BeNumerically(">=", 1))
oldCdName := cdArray[0]
exutil.By("Step4 : wait 7 mins")
//hibernateAfter is 5 min, wait for > 5 min (Need wait hardcode timer instead of check any condition here so use Sleep function directly) to check the cluster is still running status.
time.Sleep((HibernateAfterTimer + DefaultTimeout) * time.Second)
exutil.By("Step6 and Step7 : Check unclaimed cluster is still in Running status after hibernateAfter time, because of runningcount=1")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterDeployment", oldCdName, "-n", oldCdName}).check(oc)
exutil.By("Step8 :Check installedTimestap and hibernating lastTransitionTime filed.")
installedTimestap, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.installedTimestamp}`))
o.Expect(err).NotTo(o.HaveOccurred())
hibernateTransition, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].lastTransitionTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
difference := hibernateTransition.Sub(installedTimestap)
e2e.Logf("Timestamp difference is %v min", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically("<=", 1))
exutil.By("Step4 : Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
claimedTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterClaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.creationTimestamp}"))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get clusterclaim timestamp,creationTimestamp is %s", claimedTimestamp)
exutil.By("Check if ClusterClaim created successfully and become running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace(), claimName}).check(oc)
exutil.By("Check if CD is Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, 2*ClusterResumeTimeout, []string{"ClusterDeployment", oldCdName, "-n", oldCdName}).check(oc)
hibernateTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", oldCdName, "-n", oldCdName, `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].lastProbeTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get hibernattimestaps, hibernateTimestamp is %s", hibernateTimestamp)
exutil.By("Step5 : Get Timestamp difference")
difference = hibernateTimestamp.Sub(claimedTimestamp)
e2e.Logf("Timestamp difference is %v mins", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically(">=", 5))
}) | |||||
test case | openshift/openshift-tests-private | fba9348c-b222-4e53-9525-f7609f98021a | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-52415-[Azure]Hive Machinepool test for autoscale [Serial] | ['"fmt"', '"io/ioutil"', '"os"', '"path/filepath"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-52415-[Azure]Hive Machinepool test for autoscale [Serial]", func() {
testCaseID := "52415"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Azure Install-Config Secret...")
installConfigSecret := azureInstallConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
name2: cdName,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
template: filepath.Join(testDataDir, "azure-install-config.yaml"),
}
exutil.By("Config Azure ClusterDeployment...")
cluster := azureClusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AzureBaseDomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: AzureRegion,
resGroup: AzureRESGroup,
azureType: AzurePublic,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create infra MachinePool ...")
inframachinepoolAzureTemp := filepath.Join(testDataDir, "machinepool-infra-azure.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAzureTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, AzureClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Patch static replicas to autoscaler")
exutil.By("OCP-52415: [Azure]Allow minReplicas autoscaling of MachinePools to be 0")
e2e.Logf("Check hive allow set minReplicas=0 without zone setting")
autoScalingMax := "3"
autoScalingMin := "0"
removeConfig := "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig := fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 2*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check hive allow set minReplicas=0 within zone setting")
cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
infra2MachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-infra2
namespace: ` + oc.Namespace() + `
spec:
autoscaling:
maxReplicas: 3
minReplicas: 0
clusterDeploymentRef:
name: ` + cdName + `
labels:
node-role.kubernetes.io: infra2
node-role.kubernetes.io/infra2: ""
name: infra2
platform:
azure:
osDisk:
diskSizeGB: 128
type: Standard_D4s_v3
zones:
- "1"
- "2"
- "3"`
var filename = testCaseID + "-machinepool-infra2.yaml"
err = ioutil.WriteFile(filename, []byte(infra2MachinepoolYaml), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("-f", filename, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 2*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
exutil.By("Check Hive supports autoscale for Azure")
patchYaml := `
spec:
scaleDown:
enabled: true
delayAfterAdd: 10s
delayAfterDelete: 10s
delayAfterFailure: 10s
unneededTime: 10s`
e2e.Logf("Add busybox in remote cluster and check machines will scale up to maxReplicas")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ClusterAutoscaler", "default", "--type", "merge", "-p", patchYaml}).check(oc)
workloadYaml := filepath.Join(testDataDir, "workload.yaml")
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "busybox", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Deployment", "busybox", "-n", "default"}).check(oc)
e2e.Logf("Check replicas will scale up to maximum value")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Delete busybox in remote cluster and check machines will scale down to minReplicas")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas will scale down to minimum value")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | ee017e64-19dc-48ce-b573-87b358033a74 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Medium-54048-[Mag] Hive to support cli-domain-from-installer-image annotation [Serial] | ['"path/filepath"', '"strings"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_azure.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:mihuang-Medium-54048-[Mag] Hive to support cli-domain-from-installer-image annotation [Serial]", func() {
testCaseID := "54048"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
//secrets can be accessed by pod in the same namespace, so copy pull-secret and azure-credentials to target namespace for the cluster
exutil.By("Copy Azure platform credentials...")
createAzureCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "azure-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := azureInstallConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Get domain from installerimage...")
clusterVersion, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterversion/version", "-o=jsonpath={.status.desired.version}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(clusterVersion).NotTo(o.BeEmpty())
originalInstallerImage, err := getPullSpec(oc, "installer", clusterVersion)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(originalInstallerImage).NotTo(o.BeEmpty())
e2e.Logf("ClusterVersion is %s, originalInstallerImage is %s", clusterVersion, originalInstallerImage)
installerDomain := strings.SplitN(originalInstallerImage, "/", 2)[0]
installerPath := strings.SplitN(originalInstallerImage, "/", 2)[1]
overrideDomain := "mihuang.io"
overrideInstallerImage := overrideDomain + "/" + installerPath
type imageOverride struct {
copyCliDomain string
installerImageOverride string
}
var scenarios = []imageOverride{
{
"true",
overrideInstallerImage,
},
{
"false",
overrideInstallerImage,
},
{
"true",
"",
},
}
for i := 0; i < len(scenarios); i++ {
func() {
//create cluster
if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=true and installerImageOverride set...")
} else if scenarios[i].copyCliDomain == "false" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=false and installerImageOverride set...")
} else if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == "" {
exutil.By("Config Azure ClusterDeployment,with hive.openshift.io/cli-domain-from-installer-image=true and installerImageOverride unset...")
}
cluster := azureClusterDeployment{
fake: "false",
copyCliDomain: scenarios[i].copyCliDomain,
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
platformType: "azure",
credRef: AzureCreds,
region: region,
resGroup: AzureRESGroup,
azureType: cloudName,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
installerImageOverride: scenarios[i].installerImageOverride,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment-azure.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
//get conditions
exutil.By("Check if provisioning pod create success")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Provisioning", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check cd .status.installerImage")
installerImage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installerImage}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
overrideImageDomain := strings.SplitN(installerImage, "/", 2)[0]
e2e.Logf("overrideImageDomain is %s", overrideImageDomain)
e2e.Logf("Check cd .status.cliImage")
cliImage, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.cliImage}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
cliImageDomain := strings.SplitN(cliImage, "/", 2)[0]
e2e.Logf("cliImageDomain is %s", cliImageDomain)
//check conditions for scenarios
if scenarios[i].copyCliDomain == "true" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Check if both cliImage and installerImage use the new domain")
o.Expect(overrideImageDomain).To(o.Equal(overrideDomain))
o.Expect(cliImageDomain).To(o.Equal(overrideDomain))
} else if scenarios[i].copyCliDomain == "false" && scenarios[i].installerImageOverride == overrideInstallerImage {
exutil.By("Check if cliImage use offical domain and installerImage use the new domain")
o.Expect(overrideImageDomain).To(o.Equal(overrideDomain))
o.Expect(cliImageDomain).To(o.Equal(installerDomain))
} else {
exutil.By("Check if both cliImage and installerImage use the offical image")
o.Expect(overrideImageDomain).To(o.Equal(installerDomain))
o.Expect(cliImageDomain).To(o.Equal(installerDomain))
}
}()
}
}) | |||||
test | openshift/openshift-tests-private | bb11fa0d-ca43-43e5-ad30-3d9556614382 | hive_aws | import (
"context"
"crypto/ecdsa"
"crypto/elliptic"
"crypto/rand"
"crypto/tls"
"encoding/json"
"fmt"
"io/ioutil"
"math"
"net"
"os"
"os/exec"
"path/filepath"
"reflect"
"regexp"
"sort"
"strconv"
"strings"
"time"
"github.com/aws/aws-sdk-go-v2/aws"
"github.com/aws/aws-sdk-go-v2/config"
"github.com/aws/aws-sdk-go-v2/credentials"
"github.com/aws/aws-sdk-go-v2/service/cloudformation"
cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"
"github.com/aws/aws-sdk-go-v2/service/ec2"
"github.com/aws/aws-sdk-go-v2/service/ec2/types"
"github.com/aws/aws-sdk-go-v2/service/iam"
"github.com/go-acme/lego/v4/certificate"
"github.com/go-acme/lego/v4/lego"
"github.com/go-acme/lego/v4/registration"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
"github.com/openshift/openshift-tests-private/test/extended/testdata"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/apimachinery/pkg/util/wait"
e2e "k8s.io/kubernetes/test/e2e/framework"
) | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | package hive
import (
"context"
"crypto/ecdsa"
"crypto/elliptic"
"crypto/rand"
"crypto/tls"
"encoding/json"
"fmt"
"io/ioutil"
"math"
"net"
"os"
"os/exec"
"path/filepath"
"reflect"
"regexp"
"sort"
"strconv"
"strings"
"time"
"github.com/aws/aws-sdk-go-v2/aws"
"github.com/aws/aws-sdk-go-v2/config"
"github.com/aws/aws-sdk-go-v2/credentials"
"github.com/aws/aws-sdk-go-v2/service/cloudformation"
cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"
"github.com/aws/aws-sdk-go-v2/service/ec2"
"github.com/aws/aws-sdk-go-v2/service/ec2/types"
"github.com/aws/aws-sdk-go-v2/service/iam"
"github.com/go-acme/lego/v4/certificate"
"github.com/go-acme/lego/v4/lego"
"github.com/go-acme/lego/v4/registration"
g "github.com/onsi/ginkgo/v2"
o "github.com/onsi/gomega"
"github.com/openshift/openshift-tests-private/test/extended/testdata"
exutil "github.com/openshift/openshift-tests-private/test/extended/util"
"github.com/openshift/openshift-tests-private/test/extended/util/architecture"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/apimachinery/pkg/util/wait"
e2e "k8s.io/kubernetes/test/e2e/framework"
)
//
// Hive test case suite for AWS
//
var _ = g.Describe("[sig-hive] Cluster_Operator hive should", func() {
defer g.GinkgoRecover()
var (
oc = exutil.NewCLI("hive", exutil.KubeConfigPath())
ns hiveNameSpace
og operatorGroup
sub subscription
hc hiveconfig
testDataDir string
testOCPImage string
region string
basedomain string
awsPartition string
isGovCloud bool
)
// Under the hood, "extended-platform-tests run" calls "extended-platform-tests run-test" on each test case separately.
// This means that all necessary initializations need to be done before every single test case,
// either globally or in a Ginkgo node similar to BeforeEach.
g.BeforeEach(func() {
// Skip ARM64 arch
architecture.SkipNonAmd64SingleArch(oc)
// Skip if running on a non-AWS platform
exutil.SkipIfPlatformTypeNot(oc, "aws")
// The test data directory is the path to the Hive test files
testDataDir = exutil.FixturePath("testdata", "cluster_operator/hive")
// Check if MCE is enabled, if not, install Hive operator
if !isMCEEnabled(oc) {
// Install Hive operator if MCE is not enabled
_, _ = installHiveOperator(oc, &ns, &og, &sub, &hc, testDataDir)
} else {
e2e.Logf("MCE is enabled, skipping Hive operator installation.")
}
// Get OCP Image for Hive testing
testOCPImage = getTestOCPImage()
// Get platform configurations
region = getRegion(oc)
basedomain = getBasedomain(oc)
isGovCloud = strings.Contains(region, "us-gov")
awsPartition = "aws"
if isGovCloud {
e2e.Logf("Running on AWS Gov cloud")
awsPartition = "aws-us-gov"
}
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "43974" | ./bin/extended-platform-tests run --timeout 40m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-43974-Claims in excess of pool size should work well [Serial]", func() {
// Settings
var (
testCaseID = "43974"
poolName = "pool-" + testCaseID
poolName2 = "pool2-" + testCaseID
imageSetName = poolName + "-imageset"
imageSetTemp = filepath.Join(testDataDir, "clusterimageset.yaml")
pool1claimName = poolName + "-claim"
pool1claimName2 = poolName + "-claim2"
pool1claimName3 = poolName + "-claim3"
pool2claimName = poolName2 + "-claim"
pool2claimName2 = poolName2 + "-claim2"
claimTemp = filepath.Join(testDataDir, "clusterclaim.yaml")
)
// Functions
var (
checkCDNumberinClusterPool = func(checkPoolName string, expectNum int) bool {
var cdNames []string
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("cd", "-A", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
for _, cdName := range strings.Split(stdout, " ") {
if strings.Contains(cdName, checkPoolName) {
cdNames = append(cdNames, cdName)
}
}
e2e.Logf("Number of cd in ClusterPool: %d, expectNum is %d", len(cdNames), expectNum)
return len(cdNames) == expectNum
}
checkCDinClusterPoolnotDeprovisioning = func(cdArray []string) bool {
for i := range cdArray {
cdName := cdArray[i]
condition := getCondition(oc, "cd", cdName, cdName, "Provisioned")
if reason, ok := condition["reason"]; !ok || reason == "Deprovisioning" {
e2e.Logf("For condition ProvisionFailed, expected reason can not be Deprovisioning, actual reason is %v, retrying ...", reason)
return false
}
}
return true
}
)
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 4,
runningCount: 0,
maxConcurrent: 4,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool"}).check(oc)
exutil.By("Check hive will create a cd")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName, 1).
Should(o.BeTrue())
exutil.By("Create claims more than pool size 1, here creating 3 clusterclaims from above pool")
pool1claim := clusterClaim{
name: pool1claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
pool1claim2 := clusterClaim{
name: pool1claimName2,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
pool1claim3 := clusterClaim{
name: pool1claimName3,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName2})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName3})
pool1claim.create(oc)
pool1claim2.create(oc)
pool1claim3.create(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName2, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName3, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
exutil.By("Check hive will create another 3 clusterdeployments for claims")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName, 4).
Should(o.BeTrue())
cdNameArray := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdNameArray), "\n")
exutil.By("Verify clusterdeployments won't be deleted before assigned to clusterclaims. Wait for cluster finishing installation and assignment, check they won't be deleted")
o.Consistently(checkCDinClusterPoolnotDeprovisioning).
WithTimeout(FakeClusterInstallTimeout * time.Second).
WithPolling(60 * time.Second).
WithArguments(cdArray).
Should(o.BeTrue())
exutil.By("Check all clusterclaims are assigned the cluster")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName2, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName3, "-o=jsonpath={.spec.namespace}"}).check(oc)
exutil.By("Test when pool .spec.size=0, it works well too, create a clusterpool with .spec.size=0")
pool2 := clusterPool{
name: poolName2,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 0,
maxSize: 4,
runningCount: 0,
maxConcurrent: 4,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName2})
pool2.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterPool"}).check(oc)
exutil.By("check there will no cd created")
o.Consistently(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName2, 0).
Should(o.BeTrue())
exutil.By("Create 2 clusterclaim from above pool2")
pool2claim := clusterClaim{
name: pool2claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName2,
template: claimTemp,
}
pool2claim2 := clusterClaim{
name: pool2claimName2,
namespace: oc.Namespace(),
clusterPoolName: poolName2,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool2claimName})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool2claimName2})
pool2claim.create(oc)
pool2claim2.create(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool2claimName, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool2claimName2, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
exutil.By("Check hive will create 2 clusterdeployments for claims")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName2, 2).
Should(o.BeTrue())
cdNameArray = getCDlistfromPool(oc, poolName2)
cdArray = strings.Split(strings.TrimSpace(cdNameArray), "\n")
exutil.By("Verify clusterdeployments won't be deleted before assigned to clusterclaims. Wait for cluster finishing installation and assignment, check they won't be deleted")
o.Consistently(checkCDinClusterPoolnotDeprovisioning).
WithTimeout(FakeClusterInstallTimeout * time.Second).
WithPolling(60 * time.Second).
WithArguments(cdArray).
Should(o.BeTrue())
exutil.By("Check all clusterclaims are assigned the cluster")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterClaim", pool2claimName, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterClaim", pool2claimName2, "-o=jsonpath={.spec.namespace}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "46016" | ./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-46016-[HiveSpec] Test HiveConfig.Spec.FailedProvisionConfig.RetryReasons [Disruptive]", func() {
// Settings
var (
testCaseID = "46016"
cdName1 = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
cdName2 = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
retryReasons = []string{"AWSVPCLimitExceeded", "S3BucketsLimitExceeded", "NoWorkerNodes", "UnknownError", "KubeAPIWaitFailed"}
)
// AWS Clients
var (
cfg = getAWSConfig(oc, region)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
TerminateVMs = func(describeInstancesOutput *ec2.DescribeInstancesOutput) error {
var instancesToTerminate []string
for _, reservation := range describeInstancesOutput.Reservations {
instancesToTerminate = append(instancesToTerminate, *reservation.Instances[0].InstanceId)
}
_, err := ec2Client.TerminateInstances(context.Background(), &ec2.TerminateInstancesInput{
InstanceIds: instancesToTerminate,
})
e2e.Logf("Terminating VMs %v", instancesToTerminate)
return err
}
)
exutil.By("Edit hiveconfig, add RetryReasons doesn't match cluster's failure")
patch := `
spec:
failedProvisionConfig:
retryReasons:
- AWSVPCLimitExceeded
- S3BucketsLimitExceeded`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/failedProvisionConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating Install-Config Secret...")
installConfigSecretName1 := cdName1 + "-install-config"
installConfigSecret := installConfig{
name1: cdName1 + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName1,
region: region,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName1,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName1,
platformType: "aws",
credRef: AWSCreds,
region: region,
imageSetRef: cdName1 + "-imageset",
installConfigSecret: installConfigSecretName1,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 2,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Getting infraID from CD...")
infraID := getInfraIDFromCDName(oc, cdName1)
exutil.By("Waiting until the Master & bootstrap VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilMasterVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{infraID + "-master-0", infraID + "-master-1", infraID + "-master-2", infraID + "-bootstrap"},
},
},
MaxResults: aws.Int32(8),
})
return err == nil && len(describeInstancesOutput.Reservations) == 4
}
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
// The stage at which provision fails is not guaranteed. Here we just make sure provision actually fails.
exutil.By("Waiting for the first provision Pod to fail...")
provisionPod1 := getProvisionPodNames(oc, cdName1, oc.Namespace())[0]
newCheck("expect", "get", asAdmin, requireNS, compare, "Failed", ok, ClusterInstallTimeout, []string{"pod", provisionPod1, "-o=jsonpath={.status.phase}"}).check(oc)
expectKeyValue := map[string]string{
"status": "True",
"reason": "FailureReasonNotRetryable",
}
waitForHRetryFailure1 := checkCondition(oc, "ClusterDeployment", cdName1, oc.Namespace(), "ProvisionStopped", expectKeyValue, "wait for cluster installment failure for FailureReasonNotRetryable")
o.Eventually(waitForHRetryFailure1).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
exutil.By("Edit hiveconfig, add RetryReasons match cluster's failure")
patch2 := `
spec:
failedProvisionConfig:
retryReasons:
- AWSVPCLimitExceeded
- S3BucketsLimitExceeded
- NoWorkerNodes
- UnknownError
- KubeAPIWaitFailed`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating Install-Config Secret...")
installConfigSecretName2 := cdName2 + "-install-config"
installConfigSecret2 := installConfig{
name1: cdName2 + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName2,
region: region,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster2 := clusterDeployment{
fake: "false",
name: cdName2,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName2,
platformType: "aws",
credRef: AWSCreds,
region: region,
imageSetRef: cdName1 + "-imageset",
installConfigSecret: installConfigSecretName2,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 2,
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecret2.name1})
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName2})
installConfigSecret2.create(oc)
cluster2.create(oc)
exutil.By("Getting infraID from CD...")
infraID = getInfraIDFromCDName(oc, cdName2)
exutil.By("Waiting until the Master & bootstrap VMs are created...")
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Make sure the ProvisionFailed condition's reason matches one of retryReasons")
waitForProvisionFailed := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName2, oc.Namespace(), "ProvisionFailed")
if reason, ok := condition["reason"]; !ok || !ContainsInStringSlice(retryReasons, reason) {
e2e.Logf("For condition ProvisionFailed, expected reason is %v, actual reason is %v, retrying ...", retryReasons, reason)
return false
}
e2e.Logf("For condition ProvisionFailed, field reason matches one of retryReasons, proceeding to the next step ...")
return true
}
o.Eventually(waitForProvisionFailed).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
exutil.By("Getting infraID from CD again...")
var infraID2 string
getInfraIDFromCD3 := func() bool {
infraID2, _, err = oc.AsAdmin().Run("get").Args("cd", cdName2, "-o=jsonpath={.spec.clusterMetadata.infraID}").Outputs()
return err == nil && strings.HasPrefix(infraID2, cdName2) && infraID != infraID2
}
o.Eventually(getInfraIDFromCD3).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
infraID = infraID2
e2e.Logf("Found infraID = %v", infraID)
// Delete the machines again to make InstallAttempts reach the limit
exutil.By("Waiting until the Master & bootstrap VMs are created again ...")
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs again ...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue2 := map[string]string{
"status": "True",
"reason": "InstallAttemptsLimitReached",
}
waitForHRetryFailure2 := checkCondition(oc, "ClusterDeployment", cdName2, oc.Namespace(), "ProvisionStopped", expectKeyValue2, "wait for cluster installment failure for InstallAttemptsLimitReached")
o.Eventually(waitForHRetryFailure2).WithTimeout(20 * time.Minute).WithPolling(20 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "59376"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-59376-Configure resources on the hive deployment pods [Disruptive]", func() {
exutil.By("Check the default spec.resources.requests.memory value of hive controller pod")
hiveControllersPod := getHivecontrollersPod(oc, HiveNamespace)
e2e.Logf("old hivecontrollers Pod is " + hiveControllersPod)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "512Mi", ok, DefaultTimeout, []string{"pods", hiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", hiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Edit hiveconfig, add deploymentConfig sections to HiveConfig's spec")
patch := `
spec:
deploymentConfig:
- deploymentName: hive-controllers
resources:
requests:
cpu: 50m
memory: 1024Mi`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/deploymentConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Hive controller pod will restart")
var newHiveControllersPod string
checkNewcontrollersPod := func() bool {
newHiveControllersPod = getHivecontrollersPod(oc, HiveNamespace)
return strings.Compare(hiveControllersPod, newHiveControllersPod) != 0
}
o.Eventually(checkNewcontrollersPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hivecontrollers Pod is " + newHiveControllersPod)
exutil.By("Check if the new deploymentConfig applied")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1Gi", ok, DefaultTimeout, []string{"pods", newHiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", newHiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Configure deploymentConfig sections with empty resources.")
patch = `
spec:
deploymentConfig:
- deploymentName: hive-controllers
- deploymentName: hive-clustersync
- deploymentName: hiveadmission`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Verify hive-clustersync and hiveadmission pods, using the same method as hive-controllers")
hiveAdmissionPod := getHiveadmissionPod(oc, HiveNamespace)
e2e.Logf("old hiveadmission Pod is " + hiveAdmissionPod)
patch = `
spec:
deploymentConfig:
- deploymentName: hive-controllers
resources:
requests:
cpu: 50m
memory: 1024Mi
- deploymentName: hive-clustersync
resources:
requests:
cpu: 30m
memory: 600Mi
- deploymentName: hiveadmission
resources:
requests:
cpu: 50m
memory: 1024Mi`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
hiveClustersyncPod := "hive-clustersync-0"
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "600Mi", ok, DefaultTimeout, []string{"pods", hiveClustersyncPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "30m", ok, DefaultTimeout, []string{"pods", hiveClustersyncPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
var newHiveAdmissionPod string
checkNewadmissionPod := func() bool {
newHiveAdmissionPod = getHiveadmissionPod(oc, HiveNamespace)
return strings.Compare(hiveAdmissionPod, newHiveAdmissionPod) != 0
}
o.Eventually(checkNewadmissionPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hiveadmission Pod is " + newHiveAdmissionPod)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1Gi", ok, DefaultTimeout, []string{"pods", newHiveAdmissionPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", newHiveAdmissionPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Edit hiveconfig, add deploymentConfig sections to HiveConfig's spec with a bogus deploymentName, hiveconfig should not edit successfully and bounce immediately on schema validation")
patch = `
spec:
deploymentConfig:
- deploymentName: hive
resources:
requests:
cpu: 50m
memory: 1024Mi`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).To(o.HaveOccurred())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "41809"|./bin/extended-platform-tests run --timeout 8m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41809-Formalize ClusterInstall Deletion Process [Disruptive]", func() {
testCaseID := "41809"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Configure hiveconfig to enable feature gate")
patch := `
spec:
featureGates:
custom:
enabled:
- AlphaAgentInstallStrategy
featureSet: Custom`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/featureGates"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create a fake cluster")
/*
make sure:
1. no need an install-config secret, remove it from file
2. no need .spec.provisioning field in ClusterDeployment CR, remove it
3. add .spec.clusterInstallRef field in ClusterDeployment CR
4. add a new FakeClusterInstall CR in the file
*/
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
fakeClusterYaml := `
apiVersion: v1
items:
- apiVersion: hive.openshift.io/v1
kind: ClusterDeployment
metadata:
name: ` + cdName + `
spec:
baseDomain: ` + AWSBaseDomain + `
clusterName: ` + cdName + `
controlPlaneConfig:
servingCertificates: {}
installed: false
platform:
aws:
credentialsSecretRef:
name: ` + AWSCreds + `
privateLink:
enabled: false
region: ` + AWSRegion + `
pullSecretRef:
name: ` + PullSecret + `
clusterInstallRef:
group: hiveinternal.openshift.io
version: v1alpha1
kind: FakeClusterInstall
name: ` + cdName + `
status: {}
- apiVersion: hiveinternal.openshift.io/v1alpha1
kind: FakeClusterInstall
metadata:
name: ` + cdName + `
spec:
imageSetRef:
name: ` + clusterImageSetName + `
clusterDeploymentRef:
name: ` + cdName + `
kind: List
metadata: {}`
var filename = tmpDir + "/" + testCaseID + "-fakecluster.yaml"
defer os.Remove(filename)
err = os.WriteFile(filename, []byte(fakeClusterYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"FakeClusterInstall", oc.Namespace(), cdName})
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, requireNS, contain, "true", ok, FakeClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check FakeClusterInstall CR, it has an ownerReferences related to the ClusterDeployment")
newCheck("expect", "get", asAdmin, requireNS, compare, cdName, ok, DefaultTimeout, []string{"FakeClusterInstall", cdName, "-o=jsonpath={.metadata.ownerReferences[0].name}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, compare, "ClusterDeployment", ok, DefaultTimeout, []string{"FakeClusterInstall", cdName, "-o=jsonpath={.metadata.ownerReferences[0].kind}"}).check(oc)
exutil.By("Delete the clusterDeployment, check the FakeClusterInstall will be deleted too")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, requireNS, contain, cdName, nok, DefaultTimeout, []string{"FakeClusterInstall"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "39180"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Low-39180-Hive MUST can modify statefulset spec.selector when given value is non-expected value even if this field is immutable [Disruptive]", func() {
testCaseID := "39180"
resourceName := "test-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Scale down hive-operator")
// this will scale up the hive-operator and do some health check
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "{\"matchLabels\":{\"control-plane\":\"clustersync\",\"controller-tools.k8s.io\":\"1.0\"}}", ok, DefaultTimeout, []string{"sts", "-n", HiveNamespace, "-o=jsonpath={.items[].spec.selector}"}).check(oc)
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pods", "hive-clustersync-0", "-n", HiveNamespace, "-o=jsonpath={.status.phase}"}).check(oc)
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, DefaultTimeout, []string{"deployment/hive-operator", "-n", HiveNamespace, "-o=jsonpath={.status.readyReplicas}"}).check(oc)
defer oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployments", "hive-operator", "--replicas=1", "-n", HiveNamespace).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployments", "hive-operator", "--replicas=0", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Delete the existing statefulset CR")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("sts", "hive-clustersync", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Prepare a wrong statefulset CR")
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + resourceName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
statefulsetYaml := `
apiVersion: apps/v1
kind: StatefulSet
metadata:
labels:
app: hive
name: hive-clustersync
namespace: hive
spec:
replicas: 1
selector:
matchLabels:
app: hive
template:
metadata:
annotations:
labels:
app: hive
spec:
containers:
- image: quay.io/openshift-hive/hive:latest`
var filename = tmpDir + "/" + testCaseID + "-statefulset-hive.yaml"
err = os.WriteFile(filename, []byte(statefulsetYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename, "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"sts", "-n", HiveNamespace}).check(oc)
exutil.By("Check sts")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "{\"matchLabels\":{\"app\":\"hive\"}}", ok, DefaultTimeout, []string{"sts", "hive-clustersync", "-n", HiveNamespace, "-o=jsonpath={.spec.selector}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "41525"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41525-Log diffs when validation rejects immutable modifications [Serial]", func() {
testCaseID := "41525"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Patch immutable fields of ClusterDeployment")
patchCDName := "test-cluster"
patchBaseDomain := "test.com"
patchRegion := "us-east-1"
patchimageSetRefName := "test-imageset"
patch := `
spec:
baseDomain: ` + patchBaseDomain + `
clusterName: ` + patchCDName + `
platform:
aws:
region: ` + patchRegion + `
provisioning:
imageSetRef:
name: ` + patchimageSetRefName
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, `--type=merge`, "-p", patch, "-n", oc.Namespace()).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.ContainSubstring("Attempted to change ClusterDeployment.Spec which is immutable"))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("ClusterName: (%s => %s)", cdName, patchCDName)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("BaseDomain: (%s => %s)", AWSBaseDomain, patchBaseDomain)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("Platform.AWS.Region: (%s => %s)", AWSRegion, patchRegion)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("Provisioning.ImageSetRef.Name: (%s => %s)", cdName+"-imageset", patchimageSetRefName)))
exutil.By("Check .spec of ClusterDeployment, the fields tried to be changed above didn't change,")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterName}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, AWSBaseDomain, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.baseDomain}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, AWSRegion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.region}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName+"-imageset", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.provisioning.imageSetRef}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "37464"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("Author:sguo-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-37464- Seperate clustersync controller from hive-controllers, meanwhile make it be able to scale up/down [Serial]", func() {
exutil.By("Check the statefulset in hive namespace")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"statefulset", "-n", HiveNamespace}).check(oc)
exutil.By("check there is a separate pod for clustersync")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-0", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Patching HiveConfig to scale up clustersync pod")
patch := `
spec:
controllersConfig:
controllers:
- config:
replicas: 2
name: clustersync`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/controllersConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check statefulset replicas scale up to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "2", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Wait for 10 min to hive next reconcile finish, then check the hive-clustersync-1 pod is still there")
time.Sleep(10 * time.Minute)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "2", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Scale down replicas to 1 again via editing hiveconfig, check it can scale down")
patch = `
spec:
controllersConfig:
controllers:
- config:
replicas: 1
name: clustersync`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check statefulset replicas scale down to 1 again,")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-0", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", nok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "43100"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-43100-Hive supports hibernating AWS cluster with spot instances [Serial]", func() {
testCaseID := "43100"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create spots instances, one with On-Demand and another with setting maxPrice")
spotMachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-spot
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: spot
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp2
type: m4.xlarge
spotMarketOptions: {}
replicas: 1`
var filename = tmpDir + "/" + testCaseID + "-machinepool-spot.yaml"
defer os.Remove(filename)
err = ioutil.WriteFile(filename, []byte(spotMachinepoolYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-spot"})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
spotMachinepool2Yaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-spot2
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: spot2
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp2
type: m4.xlarge
spotMarketOptions:
maxPrice: "0.1"
replicas: 1`
var filename2 = tmpDir + "/" + testCaseID + "-machinepool-spot2.yaml"
defer os.Remove(filename2)
err = ioutil.WriteFile(filename2, []byte(spotMachinepool2Yaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-spot2"})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Login to target cluster, check spot instances are created")
e2e.Logf("Extracting kubeconfig ...")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
var oldSpotMachineName, oldSpotMachineName2 string
checkSpotMachineName := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("machine list: %s", stdout)
oldSpotMachineName = ""
oldSpotMachineName2 = ""
for _, MachineName := range strings.Split(stdout, " ") {
if strings.Contains(MachineName, "spot-") {
oldSpotMachineName = MachineName
}
if strings.Contains(MachineName, "spot2-") {
oldSpotMachineName2 = MachineName
}
}
e2e.Logf("oldSpotMachineName: %s, oldSpotMachineName2: %s", oldSpotMachineName, oldSpotMachineName2)
return strings.Contains(oldSpotMachineName, "spot-") && strings.Contains(oldSpotMachineName2, "spot2-")
}
o.Eventually(checkSpotMachineName).WithTimeout(DefaultTimeout * time.Second).WithPolling(5 * time.Second).Should(o.BeTrue())
// Get AWS client
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
e2e.Logf("Waiting until the spot VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilSpotVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{oldSpotMachineName, oldSpotMachineName2},
},
},
MaxResults: aws.Int32(6),
})
if err != nil {
e2e.Logf("Error when get describeInstancesOutput: %s", err.Error())
return false
}
e2e.Logf("Check result length: %d", len(describeInstancesOutput.Reservations))
for _, reservation := range describeInstancesOutput.Reservations {
instanceLen := len(reservation.Instances)
if instanceLen != 1 {
e2e.Logf("instanceLen should be 1, actual number is %d", instanceLen)
return false
}
e2e.Logf("Instance ID: %s, status: %s", *reservation.Instances[0].InstanceId, reservation.Instances[0].State.Name)
if reservation.Instances[0].State.Name != "running" {
e2e.Logf("Instances state should be running, actual state is %s", reservation.Instances[0].State.Name)
return false
}
}
return len(describeInstancesOutput.Reservations) == 2
}
o.Eventually(waitUntilSpotVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Hibernating the cluster and check ClusterDeployment Hibernating condition")
// the MachinePool can not be deleted when the ClusterDeployment is in Hibernating state
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue := map[string]string{
"status": "True",
"reason": "Hibernating",
"message": "Cluster is stopped",
}
waitForHibernating := checkCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "Hibernating", expectKeyValue, "wait for cluster hibernating")
o.Eventually(waitForHibernating).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
exutil.By("Check spot instances are terminated")
waitUntilSpotVMTerminated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{oldSpotMachineName, oldSpotMachineName2},
},
},
MaxResults: aws.Int32(6),
})
if err != nil {
e2e.Logf("Error when get describeInstancesOutput: %s", err.Error())
return false
}
e2e.Logf("Check result length: %d", len(describeInstancesOutput.Reservations))
for _, reservation := range describeInstancesOutput.Reservations {
instanceLen := len(reservation.Instances)
if instanceLen != 1 {
e2e.Logf("instanceLen should be 1, actual number is %d", instanceLen)
return false
}
e2e.Logf("Instance ID: %s, status: %s", *reservation.Instances[0].InstanceId, reservation.Instances[0].State.Name)
if reservation.Instances[0].State.Name != "terminated" {
e2e.Logf("Instances state should be terminated, actual state is %s", reservation.Instances[0].State.Name)
return false
}
}
return true
}
o.Eventually(waitUntilSpotVMTerminated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Start cluster again, check ClusterDeployment back to running again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue2 := map[string]string{
"status": "False",
"reason": "ResumingOrRunning",
"message": "Cluster is resuming or running, see Ready condition for details",
}
waitForHibernating2 := checkCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "Hibernating", expectKeyValue2, "wait for cluster being resumed")
o.Eventually(waitForHibernating2).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
e2e.Logf("Making sure the cluster is in the \"Running\" powerstate ...")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Login to target cluster, check the new spot instances are created")
var newSpotMachineName, newSpotMachineName2 string
checkSpotMachineName2 := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("machine list: %s", stdout)
newSpotMachineName = ""
newSpotMachineName2 = ""
for _, MachineName := range strings.Split(stdout, " ") {
if strings.Contains(MachineName, "spot-") {
newSpotMachineName = MachineName
}
if strings.Contains(MachineName, "spot2-") {
newSpotMachineName2 = MachineName
}
}
e2e.Logf("newSpotMachineName: %s, newSpotMachineName2: %s", newSpotMachineName, newSpotMachineName2)
return strings.Contains(newSpotMachineName, "spot-") && strings.Contains(newSpotMachineName2, "spot2-") && oldSpotMachineName != newSpotMachineName && oldSpotMachineName2 != newSpotMachineName2
}
o.Eventually(checkSpotMachineName2).WithTimeout(DefaultTimeout * time.Second).WithPolling(5 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "32135"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-32135-kubeconfig and password secrets need to be owned by ClusterDeployment after installed [Serial]", func() {
testCaseID := "32135"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check ownerReference for secrets kubeconfig and password, before installed, it is only owned by ClusterProvision.")
ClusterprovisionName := getClusterprovisionName(oc, cdName, oc.Namespace())
kubeconfigName := ClusterprovisionName + "-admin-kubeconfig"
passwordName := ClusterprovisionName + "-admin-password"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", nok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", nok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check ownership again, it will be owned by both ClusterProvision and ClusterDeployment.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Delete ClusterProvision.")
err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterProvision", ClusterprovisionName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check kubeconfig and password secrets are still exist and owned by clusterdeployment.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", nok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", nok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Delete clusterdeployment, kubeconfig and password secrets will be deleted.")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, kubeconfigName, nok, DefaultTimeout, []string{"secret", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, passwordName, nok, DefaultTimeout, []string{"secret", "-n", oc.Namespace()}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "43029"|./bin/extended-platform-tests run --timeout 20m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-43029-Hive should abandon deprovision when preserveOnDelete is true when clusters with managed DNS [Serial]", func() {
testCaseID := "43029"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create Route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, FakeClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Edit secret aws-creds and change the data to an invalid value")
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "--type", `merge`, `--patch={"data": {"aws_access_key_id": "MTIzNDU2"}}`, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Delete the cd, and then hive will hit DeprovisionLaunchError=AuthenticationFailed, and stuck in deprovision process")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
waitForDeprovisionLaunchError := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "DeprovisionLaunchError")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition DeprovisionLaunchError, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "AuthenticationFailed" {
e2e.Logf("For condition DeprovisionLaunchError, expected reason is AuthenticationFailed, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "Credential check failed") != 0 {
e2e.Logf("For condition DeprovisionLaunchError, expected message is \nCredential check failed, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition DeprovisionLaunchError, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForDeprovisionLaunchError).WithTimeout(ClusterUninstallTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
exutil.By("Set cd.spec.preserveOnDelete = true on cd")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "--type", "json", "-p", "[{\"op\": \"remove\", \"path\": \"/spec/preserveOnDelete\"}]").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "--type", `merge`, `--patch={"spec": {"preserveOnDelete": true}}`, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check in this situation, hive would be able to remove dnszone and CD CR directly")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "28631"|./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Critical-28631-[HiveSpec] Hive deprovision controller can be disabled through a hiveconfig option [Serial]", func() {
testCaseID := "28631"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
e2e.Logf("Wait until infra id generated")
newCheck("expect", "get", asAdmin, false, contain, cdName+"-", ok, 600, []string{"cd", "-n", oc.Namespace()}).check(oc)
oldhivecontrollersPod := getHivecontrollersPod(oc, HiveNamespace)
e2e.Logf("old hivecontrollers Pod is " + oldhivecontrollersPod)
e2e.Logf("Add \"deprovisionsDisabled: true\" in hiveconfig.spec")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", "json", "-p", "[{\"op\": \"remove\", \"path\": \"/spec/deprovisionsDisabled\"}]").Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"deprovisionsDisabled": true}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check \"deprovisionsDisabled\" is set to true in hiveconfig.spec")
newCheck("expect", "get", asAdmin, false, compare, "true", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.deprovisionsDisabled}"}).check(oc)
e2e.Logf("Check if hivecontrollers Pod is recreated")
var hivecontrollersPod string
checkNewcontrollersPod := func() bool {
hivecontrollersPod = getHivecontrollersPod(oc, HiveNamespace)
return strings.Compare(oldhivecontrollersPod, hivecontrollersPod) != 0
}
o.Eventually(checkNewcontrollersPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hivecontrollers Pod is " + hivecontrollersPod)
e2e.Logf("Try to delete cd")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
e2e.Logf(`Check logs of hive-controllers has a warning :"deprovisions are currently disabled in HiveConfig, skipping"`)
checkDeprovisionLog := func() bool {
deprovisionLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hivecontrollersPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(deprovisionLogs, "deprovisions are currently disabled in HiveConfig, skipping") {
e2e.Logf(`Find target message :"deprovisions are currently disabled in HiveConfig, skipping"`)
return true
}
e2e.Logf(`Still waiting for message :"deprovisions are currently disabled in HiveConfig, skipping"`)
return false
}
o.Eventually(checkDeprovisionLog).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
e2e.Logf("Add \"deprovisionsDisabled: false\" in hiveconfig.spec")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"deprovisionsDisabled": false}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check \"deprovisionsDisabled\" is set to false in hiveconfig.spec")
newCheck("expect", "get", asAdmin, false, compare, "false", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.deprovisionsDisabled}"}).check(oc)
e2e.Logf("Check if cd is in deprovision.")
newCheck("expect", "get", asAdmin, false, contain, cdName+"-uninstall-", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace()}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "41286"|./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-41286-ClusterPool supports provisioning fake cluster [Serial]", func() {
testCaseID := "41286"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
exutil.By("Check hive will propagate the annotation to all created ClusterDeployment")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
newCheck("expect", "get", asAdmin, withoutNamespace, contain, `"hive.openshift.io/fake-cluster":"true"`, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.metadata.annotations}"}).check(oc)
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check claiming a fake cluster works well")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
waitForClusterClaimRunning := func() bool {
condition := getCondition(oc, "ClusterClaim", claimName, oc.Namespace(), "ClusterRunning")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition ClusterRunning, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Running" {
e2e.Logf("For condition ClusterRunning, expected reason is Running, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || message != "Cluster is running" {
e2e.Logf("For condition ClusterRunning, expected message is \nCluster is running, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition ClusterRunning, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClusterClaimRunning).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check clusterMetadata field of fake cluster, all fields have values")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.adminKubeconfigSecretRef.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.adminPasswordSecretRef.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.infraID}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "42661"|./bin/extended-platform-tests run --timeout 25m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-42661-Low-39179-Simulate hibernation for fake clusters [Serial]", func() {
testCaseID := "42661"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create fake ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "1m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Check all clusters in cluster pool are in Hibernating status")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check claiming a fake cluster works well")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check cluster is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Hibernating it again, check it can be hibernated again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", cdNameSpace, "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create fake ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
installConfigSecret.create(oc)
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("Check fake cluster is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Hibernating the fake cluster ,check it can be hibernated")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Restart it again, check it back to running again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("OCP-39179: Autogenerate Cluster ID in hive for fake cluster instead of fixed value fake-cluster-id")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "fake-cluster-id", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "fake-cluster-", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
ClusterprovisionName := getClusterprovisionName(oc, cdName, oc.Namespace())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "fake-cluster-id", nok, DefaultTimeout, []string{"ClusterProvision", ClusterprovisionName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "fake-cluster-", ok, DefaultTimeout, []string{"ClusterProvision", ClusterprovisionName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterID}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "25443"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Low-25443-Low-29855-Clusterdeployment contains Status.Condition of SyncSet status in case of syncset is invalid [Serial]", func() {
testCaseID := "25443"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create SyncSet for resource apply......")
syncSetName := testCaseID + "-syncset1"
configMapName := testCaseID + "-configmap1"
configMapNamespace := testCaseID + "-configmap1-ns"
resourceMode := "Sync"
syncTemp := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource := syncSetResource{
name: syncSetName,
namespace: oc.Namespace(),
namespace2: configMapNamespace,
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: resourceMode,
template: syncTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName})
syncResource.create(oc)
e2e.Logf("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Check if SyncSetPatch is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName, ok, DefaultTimeout, []string{"SyncSet", syncSetName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if Syncset is not failed before applying the patch.")
waitForSyncsetSuccess := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "SyncSetFailed")
if status, ok := condition["status"]; !ok || status != "False" {
e2e.Logf("For condition SyncSetFailed, expected status is False, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "SyncSetApplySuccess" {
e2e.Logf("For condition SyncSetFailed, expected reason is SyncSetApplySuccess, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "SyncSet apply is successful") != 0 {
e2e.Logf("For condition SyncSetFailed, expected message is \nSyncSet apply is successful, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition SyncSetFailed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForSyncsetSuccess).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
syncSetPatchName := testCaseID + "-syncset-patch"
syncPatchTemp := filepath.Join(testDataDir, "syncset-patch.yaml")
patchContent := ` { "data": { "foo": "new-bar" }`
patchType := "merge"
syncPatch := syncSetPatch{
name: syncSetPatchName,
namespace: oc.Namespace(),
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
pcontent: patchContent,
patchType: patchType,
template: syncPatchTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
syncPatch.create(oc)
e2e.Logf("Check if Syncset is failed.")
waitForSyncsetFail := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "SyncSetFailed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition SyncSetFailed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "SyncSetApplyFailure" {
e2e.Logf("For condition SyncSetFailed, expected reason is SyncSetApplyFailure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "One of the SyncSet applies has failed") != 0 {
e2e.Logf("For condition SyncSetFailed, expected message is \nOne of the SyncSet applies has failed, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition SyncSetFailed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForSyncsetFail).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-29855:Hive treates bad syncsets as controller warnings instead of controller errors")
waitForClustersyncFail1 := func() bool {
condition := getCondition(oc, "clustersync", cdName, oc.Namespace(), "Failed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition Failed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Failure" {
e2e.Logf("For condition Failed, expected reason is Failure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, fmt.Sprintf("SyncSet %s is failing", syncSetPatchName)) != 0 {
e2e.Logf("For condition Failed, expected message is \nSyncSet %v is failing, \nactual reason is %v\n, retrying ...", syncSetPatchName, message)
return false
}
e2e.Logf("For condition Failed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClustersyncFail1).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
hiveclustersyncPod := "hive-clustersync-0"
e2e.Logf(`Check logs of hive-clustersync-0 has a warning log instead of error log`)
checkclustersyncLog1 := func() bool {
clustersyncLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveclustersyncPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(clustersyncLogs, "level=warning msg=\"running the patch command failed\"") {
e2e.Logf(`Find target message :level=warning msg="running the patch command failed"`)
return true
}
e2e.Logf(`Still waiting for message :level=warning msg="running the patch command failed"`)
return false
}
o.Eventually(checkclustersyncLog1).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
exutil.By("Extracting kubeconfig ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
syncSetSecretName := testCaseID + "-syncset-secret"
syncSecretTemp := filepath.Join(testDataDir, "syncset-secret.yaml")
sourceName := testCaseID + "-secret"
syncSecret := syncSetSecret{
name: syncSetSecretName,
namespace: oc.Namespace(),
cdrefname: cdName,
sname: "secret-not-exist",
snamespace: oc.Namespace(),
tname: sourceName,
tnamespace: "default",
template: syncSecretTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetSecretName})
syncSecret.create(oc)
e2e.Logf("Check if Syncset-secret failed to apply.")
waitForClustersyncFail2 := func() bool {
condition := getCondition(oc, "clustersync", cdName, oc.Namespace(), "Failed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition Failed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Failure" {
e2e.Logf("For condition Failed, expected reason is Failure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, fmt.Sprintf("SyncSet %s is failing", syncSetSecretName)) != 0 {
e2e.Logf("For condition Failed, expected message is \nSyncSet %v is failing, \nactual reason is %v\n, retrying ...", syncSetSecretName, message)
return false
}
e2e.Logf("For condition Failed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClustersyncFail2).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("Check target cluster doesn't have this secret.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "secret"}).check(oc)
e2e.Logf(`Check logs of hive-clustersync-0 doesn't have error log`)
checkclustersyncLog2 := func() bool {
clustersyncLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveclustersyncPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(clustersyncLogs, fmt.Sprintf("level=info msg=\"cannot read secret\" SyncSet=%s", syncSetSecretName)) {
e2e.Logf(`Find target message :level=info msg="cannot read secret"`)
return true
}
e2e.Logf(`Still waiting for message :level=info msg="cannot read secret"`)
return false
}
o.Eventually(checkclustersyncLog2).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "23986"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Critical-23986-Medium-64550-[HiveSpec] Kubeconfig secrets can work with additional CAs[Serial]", func() {
testCaseID := "23986"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
apiEndpoint := "api." + cdName + "." + AWSBaseDomain
appsEndpoint := "apps." + cdName + "." + AWSBaseDomain
appsEndpointGlobbing := "*." + appsEndpoint
appsEndpointConsole := "console-openshift-console." + appsEndpoint
/*
To generate a Let's Encrypt certificate, we have the following options:
1) Use the cert-manager operator:
Pro: Openshift native
Con: we are no longer testing Hive itself as we rely on another operator as well
2) Use certbot (or hiveutil which relies on it):
Pro: straightforwardness
Con: we have to install certbot
3) Use a Golang library which automates this process:
Pro: straightforwardness (somewhat)
Con: cannot think of any
Here we are using option 3).
*/
exutil.By("Getting a Let's Encrypt certificate for " + apiEndpoint + " & " + appsEndpointGlobbing)
// Get Lego user and config
privateKey, err := ecdsa.GenerateKey(elliptic.P256(), rand.Reader)
o.Expect(err).NotTo(o.HaveOccurred())
user := legoUser{key: privateKey}
config := lego.NewConfig(&user)
// Get Lego client
client, err := lego.NewClient(config)
o.Expect(err).NotTo(o.HaveOccurred())
// Registration for new user
_, err = client.Registration.Register(registration.RegisterOptions{TermsOfServiceAgreed: true})
o.Expect(err).NotTo(o.HaveOccurred())
// Set Lego DNS provider which is used to solve the ACME DNS challenge
// (and cleanup the related DNS records after that)
maxRetries := 5
TTL := 10
propagationTimeout, pollingInterval := 15*time.Minute, 4*time.Second
awsAccessKeyId, awsSecretAccessKey := getAWSCredentials(oc)
dnsProvider, err := newLegoDNSProvider(maxRetries, TTL, propagationTimeout, pollingInterval, awsAccessKeyId, awsSecretAccessKey, AWSRegion)
o.Expect(err).NotTo(o.HaveOccurred())
err = client.Challenge.SetDNS01Provider(dnsProvider)
o.Expect(err).NotTo(o.HaveOccurred())
// Request for certificates
// Note:
// Lego checks DNS record propagation from recursive DNS servers specified in /etc/resolv.conf (if possible).
// So before running this test case locally, turn off the VPNs as they often update /etc/resolv.conf.
request := certificate.ObtainRequest{
Domains: []string{apiEndpoint, appsEndpointGlobbing},
// We want the certificates to be split
Bundle: false,
}
certificates, err := client.Certificate.Obtain(request)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Writing certificates & private key to files...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
fullChainFilePath := tmpDir + "/fullchain.pem"
err = os.WriteFile(fullChainFilePath, append(certificates.Certificate, certificates.IssuerCertificate...), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
chainFilePath := tmpDir + "/chain.pem"
err = os.WriteFile(chainFilePath, certificates.IssuerCertificate, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
privateKeyFilePath := tmpDir + "/privkey.pem"
err = os.WriteFile(privateKeyFilePath, certificates.PrivateKey, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating serving-cert Secret which will be referenced in CD's manifest...")
servingCertificateSecretName := "serving-cert"
defer oc.AsAdmin().Run("delete").Args("secret", servingCertificateSecretName).Execute()
err = oc.AsAdmin().Run("create").Args("secret", "tls", servingCertificateSecretName, "--cert="+fullChainFilePath, "--key="+privateKeyFilePath).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ca-cert Secret which will be referenced in HiveConfig/hive...")
caCertificateSecretName := "ca-cert"
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", caCertificateSecretName, "-n=hive").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", caCertificateSecretName, "--from-file=ca.crt="+chainFilePath, "-n=hive").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Referencing ca-cert Secret in HiveConfig/hive...")
patch := `
spec:
additionalCertificateAuthoritiesSecretRef:
- name: ` + caCertificateSecretName
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/additionalCertificateAuthoritiesSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterDeployment...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cd := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, cd.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cd.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cd)
exutil.By("Patching CD s.t. it references the serving certificate Secret...")
patch = fmt.Sprintf(`
spec:
certificateBundles:
- name: serving-cert
certificateSecretRef:
name: %s
controlPlaneConfig:
servingCertificates:
default: serving-cert
ingress:
- name: default
domain: %s
servingCertificate: serving-cert`, servingCertificateSecretName, appsEndpoint)
err = oc.AsAdmin().Run("patch").Args("clusterdeployment", cdName, "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Waiting for the CD to be installed...")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Making sure the target cluster is using the right certificate...")
endpointCertIsGood := func(endpoint string) bool {
e2e.Logf("Checking certificates for endpoint %v ...", endpoint)
conn, err := tls.Dial("tcp", endpoint, &tls.Config{InsecureSkipVerify: true})
if err != nil {
e2e.Logf("Error dialing endpoint %v: %v, keep polling ...", endpoint, err.Error())
return false
}
// Must call conn.Close() here to make sure the connection is successfully established,
// so the conn object is populated and can be closed without incurring a nil pointer dereference error.
defer conn.Close()
// Look for the target certificate (the one with apiEndpoint/appsEndpoint as subject)
// in all certificates of the endpoint
for _, cert := range conn.ConnectionState().PeerCertificates {
if strings.Contains(cert.Subject.String(), apiEndpoint) || strings.Contains(cert.Subject.String(), appsEndpoint) {
// For simplicity, here we only check the issuer is correct on the target certificate
return strings.Contains(cert.Issuer.String(), `Let's Encrypt`)
}
}
e2e.Logf("Target certificate not found on endpoint %v, keep polling ...", endpoint)
return false
}
// It seems that DNS propagation can be really slow for "*.apps.CLUSTER.qe.devcluster.openshift.com" (literally)
// So here we check the console endpoint "console.apps.CLUSTER.qe.devcluster.openshift.com" instead
checkCertificates := func() bool {
return endpointCertIsGood(apiEndpoint+":6443") && endpointCertIsGood(appsEndpointConsole+":443")
}
// We need to poll s.t. remote-ingress or control-plane-certificate-related SyncSets are applied
// and APIServer/Ingress-Operator finish reconcile on the target cluster.
o.Eventually(checkCertificates).WithTimeout(20 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
// The kubeconfig obtained (for ex. Secret/fxie-hive-1-0-wlqg2-admin-kubeconfig.data["kubeconfig"]) has the
// CA certs integrated, so we should be able to communicate to the target cluster without the following error:
// "x509: certificate signed by unknown authority".
exutil.By("Communicating to the target cluster using the kubeconfig with Let's Encrypt's CA...")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfigPath := tmpDir + "/kubeconfig"
err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "--kubeconfig", kubeconfigPath).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("OCP-64550: Hive should be able to delete Secret/hive-additional-ca")
// Make sure the hive-additional-CA Secret still exists at this moment
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", hiveAdditionalCASecret, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).To(o.ContainSubstring(hiveAdditionalCASecret))
// Patch HiveConfig
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/additionalCertificateAuthoritiesSecretRef"}]`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
// Make sure the hive-additional-CA Secret is eventually deleted
hiveOperatorReconcileTimeout := 300
newCheck("expect", "get", asAdmin, withoutNamespace, contain, hiveAdditionalCASecret, nok, hiveOperatorReconcileTimeout, []string{"Secret", "-n", HiveNamespace}).check(oc)
// Make sure Hive Operator stays healthy for a while
hiveIsStillHealthy := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("hiveconfig/hive", `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return stdout == "True"
}
o.Consistently(hiveIsStillHealthy).WithTimeout(DefaultTimeout * time.Second).WithPolling(10 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "25145"|./bin/extended-platform-tests run --timeout 45m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25145-[HiveSpec] Dynamically detect change to global pull secret content [Serial]", func() {
testCaseID := "25145"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Preparing an incomplete pull-secret ...")
var pullSecretMapIncomplete map[string]map[string]map[string]string
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/pull-secret", "-n", "openshift-config", "--to", "-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &pullSecretMapIncomplete)
o.Expect(err).NotTo(o.HaveOccurred())
delete(pullSecretMapIncomplete["auths"], "registry.ci.openshift.org")
exutil.By("Creating an incomplete pull-secret in Hive's namespace and the temporary project's namespace respectively ...")
pullSecretBsIncomplete, _ := json.Marshal(pullSecretMapIncomplete)
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", PullSecret, "-n", HiveNamespace).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", PullSecret, "--from-literal=.dockerconfigjson="+string(pullSecretBsIncomplete), "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.Run("delete").Args("secret", PullSecret).Execute()
err = oc.Run("create").Args("secret", "generic", PullSecret, "--from-literal=.dockerconfigjson="+string(pullSecretBsIncomplete)).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Patching HiveConfig so that it refers to an incomplete global pull-secret ...")
patch := `
spec:
globalPullSecretRef:
name: ` + PullSecret
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/globalPullSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterImageSet ...")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret ...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Copying AWS credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment with an incomplete pull-secret ...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("Waiting for the cluster installation to fail ...")
waitForAPIWaitFailure := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "ProvisionFailed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition ProvisionFailed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "KubeAPIWaitFailed" {
e2e.Logf("For condition ProvisionFailed, expected reason is KubeAPIWaitFailed, actual reason is %v, retrying ...", reason)
return false
}
e2e.Logf("For condition ProvisionFailed, fields status and reason expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForAPIWaitFailure).WithTimeout(ClusterInstallTimeout * time.Second).WithPolling(3 * time.Minute).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "25210"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25210-Collect ClusterOperator Status for Hive Managed Clusters [Serial]", func() {
testCaseID := "25210"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating install-config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Making sure the cluster is installed and in the \"Running\" powerstate ...")
newCheck("expect", "get", asAdmin, false, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, false, compare, "Running", ok, WaitingForClusterOperatorsTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Extracting kubeconfig ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
exutil.By("Comparing conditions obtained from ClusterOperator and ClusterState ...")
var clusterStateConditions, clusterOperatorConditions map[string][]map[string]string
clusterStateJSONPath := `{"{"}{range .status.clusterOperators[:-1]}"{.name}":{.conditions},{end}{range .status.clusterOperators[-1]}"{.name}":{.conditions}{end}{"}"}`
clusterOperatorJSONPath := `{"{"}{range .items[:-1]}"{.metadata.name}":{.status.conditions},{end}{range .items[-1]}"{.metadata.name}":{.status.conditions}{end}{"}"}`
/*
stdout[any-index] =
{
"operator-name": [
{
"lastTransitionTime": ...
...
}
]
}
*/
checkConditionEquality := func() bool {
stdout, _, err := oc.AsAdmin().Run("get").Args("ClusterState", cdName, "-o", "jsonpath="+clusterStateJSONPath).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &clusterStateConditions)
o.Expect(err).NotTo(o.HaveOccurred())
stdout, _, err = oc.AsAdmin().Run("get").Args("ClusterOperator", "-o", "jsonpath="+clusterOperatorJSONPath, "--kubeconfig="+kubeconfig).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &clusterOperatorConditions)
o.Expect(err).NotTo(o.HaveOccurred())
return reflect.DeepEqual(clusterOperatorConditions, clusterStateConditions)
}
o.Eventually(checkConditionEquality).WithTimeout(20 * time.Minute).WithPolling(time.Minute).Should(o.BeTrue())
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "33832"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-33832-Low-42251-Medium-43033-[HiveSpec] Hive supports ClusterPool [Serial]", func() {
testCaseID := "33832"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-42251 - Initialize hive CR conditions")
exutil.By("OCP-42251 Step 1: Check all conditions type of ClusterPool")
allClusterPoolConditionTypes := []string{"MissingDependencies", "CapacityAvailable", "AllClustersCurrent", "InventoryValid", "DeletionPossible"}
sort.Strings(allClusterPoolConditionTypes)
checkClusterPoolConditionType := func() bool {
checkedClusterPoolConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterPoolConditionTypes := strings.Split(checkedClusterPoolConditionTypesOutput, " ")
sort.Strings(checkedClusterPoolConditionTypes)
e2e.Logf("Compare allClusterPoolConditionTypes: %v and checkedClusterPoolConditionTypes: %v", allClusterPoolConditionTypes, checkedClusterPoolConditionTypes)
return reflect.DeepEqual(allClusterPoolConditionTypes, checkedClusterPoolConditionTypes)
}
o.Eventually(checkClusterPoolConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("Check if ClusterDeployment is created")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterDeployment", "-A"}).check(oc)
exutil.By("OCP-42251 Step 2: Check all conditions type of ClusterDeployment")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
allClusterDeploymentConditionTypes := []string{"InstallerImageResolutionFailed", "ControlPlaneCertificateNotFound", "IngressCertificateNotFound", "Unreachable", "ActiveAPIURLOverride",
"DNSNotReady", "InstallImagesNotResolved", "ProvisionFailed", "SyncSetFailed", "RelocationFailed", "Hibernating", "Ready", "InstallLaunchError", "DeprovisionLaunchError", "ProvisionStopped",
"Provisioned", "RequirementsMet", "AuthenticationFailure", "AWSPrivateLinkReady", "AWSPrivateLinkFailed", "ClusterInstallFailed", "ClusterInstallCompleted", "ClusterInstallStopped", "ClusterInstallRequirementsMet"}
sort.Strings(allClusterDeploymentConditionTypes)
checkClusterDeploymentConditionType := func() bool {
checkedClusterDeploymentConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterDeploymentConditionTypes := strings.Split(checkedClusterDeploymentConditionTypesOutput, " ")
sort.Strings(checkedClusterDeploymentConditionTypes)
e2e.Logf("Compare allClusterDeploymentConditionTypes: %v and checkedClusterDeploymentConditionTypes: %v", allClusterDeploymentConditionTypes, checkedClusterDeploymentConditionTypes)
return reflect.DeepEqual(allClusterDeploymentConditionTypes, checkedClusterDeploymentConditionTypes)
}
o.Eventually(checkClusterDeploymentConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-42251 Step 3: Check all conditions type of MachinePool")
machinepoolName := cdName + "-worker"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinepoolName, ok, DefaultTimeout, []string{"Machinepool", "-n", cdNameSpace}).check(oc)
allMachinepoolConditionTypes := []string{"NotEnoughReplicas", "NoMachinePoolNameLeasesAvailable", "InvalidSubnets", "UnsupportedConfiguration"}
sort.Strings(allMachinepoolConditionTypes)
checkMachinePoolConditionType := func() bool {
checkedMachinepoolConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Machinepool", machinepoolName, "-n", cdNameSpace, "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedMachinepoolConditionTypes := strings.Split(checkedMachinepoolConditionTypesOutput, " ")
sort.Strings(checkedMachinepoolConditionTypes)
e2e.Logf("Compare allMachinepoolConditionTypes: %v and checkedMachinepoolConditionTypes: %v", allMachinepoolConditionTypes, checkedMachinepoolConditionTypes)
return reflect.DeepEqual(allMachinepoolConditionTypes, checkedMachinepoolConditionTypes)
}
o.Eventually(checkMachinePoolConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check if ClusterPool become ready")
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, ClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-42251 Step 4: Check all conditions type of ClusterClaim")
allClusterClaimConditionTypes := []string{"Pending", "ClusterRunning"}
sort.Strings(allClusterClaimConditionTypes)
checkClusterClaimConditionType := func() bool {
checkedClusterClaimConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterClaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterClaimConditionTypes := strings.Split(checkedClusterClaimConditionTypesOutput, " ")
sort.Strings(checkedClusterClaimConditionTypes)
e2e.Logf("Compare allClusterClaimConditionTypes: %v and checkedClusterClaimConditionTypes: %v", allClusterClaimConditionTypes, checkedClusterClaimConditionTypes)
return reflect.DeepEqual(allClusterClaimConditionTypes, checkedClusterClaimConditionTypes)
}
o.Eventually(checkClusterClaimConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check if ClusterClaim become running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-43033: oc get clusterclaim should report ClusterDeleted")
exutil.By("Delete the ClusterDeployment")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", cdNameSpace).Background()
defer cmd.Process.Kill()
exutil.By("Check ClusterRunning conditions of clusterclaim")
expectKeyValue := map[string]string{
"status": "False",
"reason": "ClusterDeleted",
"message": "Assigned cluster has been deleted",
}
waitForClusterRunningFalse := checkCondition(oc, "ClusterClaim", claimName, oc.Namespace(), "ClusterRunning", expectKeyValue, "wait for ClusterRunning false")
o.Eventually(waitForClusterRunningFalse).WithTimeout(ClusterUninstallTimeout * time.Second).WithPolling(15 * time.Second).Should(o.BeTrue())
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "23167"|./bin/extended-platform-tests run --timeout 55m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-23167-Critical-73287-The tags created on users in AWS match what the installer did on your instances [Serial]", func() {
testCaseID := "23167"
cdName := "cd-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating ClusterDeployment ...")
installConfig := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cd := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cd.name+"-imageset", oc.Namespace(), installConfig.name1, cd.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfig, cd)
exutil.By("Create the infra MachinePool ...")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
inframp.create(oc)
// Wait for the cluster to be installed and extract its infra id
newCheck("expect", "get", asAdmin, false, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
infraID, _, err := oc.AsAdmin().Run("get").Args("cd", cdName, "-o", "jsonpath='{.spec.clusterMetadata.infraID}'").Outputs()
infraID = strings.Trim(infraID, "'")
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Cluster infraID: " + infraID)
// Extract AWS credentials
AWSAccessKeyID, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/aws-creds", "-n=kube-system", "--keys=aws_access_key_id", "--to=-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
AWSSecretAccessKey, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/aws-creds", "-n=kube-system", "--keys=aws_secret_access_key", "--to=-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// AWS clients
cfg, err := config.LoadDefaultConfig(
context.Background(),
config.WithCredentialsProvider(credentials.NewStaticCredentialsProvider(AWSAccessKeyID, AWSSecretAccessKey, "")),
config.WithRegion(AWSRegion),
)
o.Expect(err).NotTo(o.HaveOccurred())
ec2Client := ec2.NewFromConfig(cfg)
iamClient := iam.NewFromConfig(cfg)
// Make sure resources are created with the target tag
targetTag := "kubernetes.io/cluster/" + infraID
exutil.By("Checking that resources are created with the target tag " + targetTag)
describeTagsOutput, err := ec2Client.DescribeTags(context.Background(), &ec2.DescribeTagsInput{
Filters: []types.Filter{
{
Name: aws.String("key"),
Values: []string{targetTag},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeTagsOutput.Tags)).NotTo(o.BeZero())
// Make sure the IAM users are tagged
exutil.By("Looking for IAM users prefixed with infraID ...")
pagination := aws.Int32(50)
userFound, username := false, ""
listUsersOutput := &iam.ListUsersOutput{}
err = wait.Poll(6*time.Second, 10*time.Minute, func() (bool, error) {
listUsersOutput, err = iamClient.ListUsers(context.Background(), &iam.ListUsersInput{
Marker: listUsersOutput.Marker,
MaxItems: pagination,
})
o.Expect(err).NotTo(o.HaveOccurred())
for _, user := range listUsersOutput.Users {
if strings.HasPrefix(*user.UserName, infraID) {
userFound, username = true, *user.UserName
break
}
}
if userFound {
return true, nil
}
return false, nil
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(userFound).To(o.BeTrue())
exutil.By("Looking for tags on user " + username)
listUserTagsOutput, err := iamClient.ListUserTags(context.Background(), &iam.ListUserTagsInput{
UserName: aws.String(username),
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(*listUserTagsOutput.Tags[0].Key).To(o.Equal(targetTag))
o.Expect(*listUserTagsOutput.Tags[0].Value).To(o.Equal("owned"))
exutil.By("Check OCP-73287, the machine in infra MP shall have userTags defined in both CD and MP")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
instanceId := getResource(oc, asAdmin, withoutNamespace, "machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[0].status.providerStatus.instanceId}", "--kubeconfig="+kubeconfig)
e2e.Logf("instanceId is %v", instanceId)
describeTagsOutput, err = ec2Client.DescribeTags(context.Background(), &ec2.DescribeTagsInput{
Filters: []types.Filter{
{
Name: aws.String("resource-id"),
Values: []string{instanceId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
tagsLen := len(describeTagsOutput.Tags)
e2e.Logf("Tags length = %v", tagsLen)
numMatchedTags := 0
for i := range describeTagsOutput.Tags {
e2e.Logf("userTag %v, Key = %v, Value = %v", i, *describeTagsOutput.Tags[i].Key, *describeTagsOutput.Tags[i].Value)
// Tag defined in CD
if *describeTagsOutput.Tags[i].Key == "hive-qe-tag1" && *describeTagsOutput.Tags[i].Value == "hive-qe-value1" {
numMatchedTags++
}
// Tag defined in MP
if *describeTagsOutput.Tags[i].Key == "hive-qe-tag3" && *describeTagsOutput.Tags[i].Value == "hive-qe-value3" {
numMatchedTags++
}
// Tag defined in both CD and MP, MP value shall take precedence
if *describeTagsOutput.Tags[i].Key == "hive-qe-customizedtag" && *describeTagsOutput.Tags[i].Value == AWSDefaultMPTag {
numMatchedTags++
}
}
//Totally 3 tags matched
o.Expect(numMatchedTags).To(o.Equal(3))
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "25310"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-25310-High-33374-High-39747-Medium-23165-High-22760- [HiveSDRosa] Hive ClusterDeployment Check installed and version [Serial]", func() {
testCaseID := "25310"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Selecting a custom OCP version to install ...")
ocpVersion := extractRelFromImg(testOCPImage)
xyzVersion := strings.Split(ocpVersion, ".")
majorVersion := xyzVersion[0]
minorVersion := xyzVersion[1]
patchVersion := xyzVersion[2]
minorVersionInt, err := strconv.Atoi(minorVersion)
o.Expect(err).NotTo(o.HaveOccurred())
minorVersion = strconv.Itoa(minorVersionInt - 1)
customOCPImage, err := exutil.GetLatestNightlyImage(majorVersion + "." + minorVersion)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Will install OCP version " + customOCPImage)
exutil.By("config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, customOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("hive.go namespace..." + oc.Namespace())
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
e2e.Logf("test OCP-25310")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("test OCP-33374")
ocpVersion = majorVersion + "." + minorVersion + "." + patchVersion
if ocpVersion == "" {
g.Fail("Case failed because no OCP version extracted from Image")
}
if ocpVersion != "" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, ocpVersion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels}"}).check(oc)
}
e2e.Logf("test OCP-39747")
if ocpVersion != "" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, ocpVersion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installVersion}"}).check(oc)
}
exutil.By("OCP-23165:Hive supports remote Machine Set Management for AWS")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check worker machinepool .status.replicas = 3")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check only 1 machineset up")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check only one machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 3")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 3}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale up to 3")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 3 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 2")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 2}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale down to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 2 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("OCP-22760: Use custom cluster image set to deploy cluster")
fullImgString := customOCPImage[strings.Index(customOCPImage, ":")+1:]
newCheck("expect", "get", asAdmin, withoutNamespace, compare, fullImgString, ok, DefaultTimeout, []string{"ClusterVersion", "version", "-o=jsonpath={.status.desired.version}", "--kubeconfig=" + kubeconfig}).check(oc)
})
//author: [email protected]
//OCP-44945, OCP-37528, OCP-37527
//example: ./bin/extended-platform-tests run all --dry-run|grep "44945"|./bin/extended-platform-tests run --timeout 35m -f -
g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-44945-Low-37528-Low-37527-[HiveSpec] Hive supports ClusterPool runningCount and hibernateAfter[Serial]", func() {
testCaseID := "44945"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
e2e.Logf("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "10m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
e2e.Logf("Check if ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 2, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
e2e.Logf("OCP-44945, step 2: check all cluster are in Hibernating status")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
e2e.Logf("OCP-37528, step 3: check hibernateAfter and powerState fields")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.spec.powerState}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10m", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.spec.hibernateAfter}"}).check(oc)
}
exutil.By("OCP-44945, step 5: Patch .spec.runningCount=1...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"runningCount":1}}`}).check(oc)
e2e.Logf("OCP-44945, step 6: Check the unclaimed clusters in the pool, CD whose creationTimestamp is the oldest becomes Running")
var oldestCD, oldestCDTimestamp string
oldestCDTimestamp = ""
for i := range cdArray {
creationTimestamp := getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.creationTimestamp}")
e2e.Logf("CD %d is %s, creationTimestamp is %s", i, cdArray[i], creationTimestamp)
if strings.Compare(oldestCDTimestamp, "") == 0 || strings.Compare(oldestCDTimestamp, creationTimestamp) > 0 {
oldestCDTimestamp = creationTimestamp
oldestCD = cdArray[i]
}
}
e2e.Logf("The CD with the oldest creationTimestamp is %s", oldestCD)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD}).check(oc)
exutil.By("OCP-44945, step 7: Patch pool.spec.runningCount=3...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"runningCount":3}}`}).check(oc)
e2e.Logf("OCP-44945, step 7: check runningCount=3 but pool size is still 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.spec.runningCount}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.spec.size}"}).check(oc)
e2e.Logf("OCP-44945, step 7: All CDs in the pool become Running")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
exutil.By("OCP-44945, step 8: Claim a CD from the pool...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
e2e.Logf("OCP-44945, step 8: Check the claimed CD is the one whose creationTimestamp is the oldest")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, oldestCD, ok, ClusterResumeTimeout, []string{"ClusterClaim", claimName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("OCP-44945, step 9: Check CD's ClaimedTimestamp is set")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "claimedTimestamp", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.clusterPoolRef}"}).check(oc)
e2e.Logf("OCP-37528, step 5: Check the claimed CD is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("OCP-37528, step 6: Check the claimed CD is in Hibernating status due to hibernateAfter=10m")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout+5*DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
exutil.By("OCP-37527, step 4: patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("OCP-37527, step 5: CD becomes Hibernating again due to hibernateAfter=10m")
//patch makes CD to be Running soon but it needs more time to get back from Hibernation actually so overall timer is ClusterResumeTimeout + hibernateAfter
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout+5*DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
})
//author: [email protected] [email protected]
//OCP-23040, OCP-42113, OCP-34719, OCP-41250, OCP-25334, OCP-23876
//example: ./bin/extended-platform-tests run all --dry-run|grep "23040"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-23040-Medium-42113-High-34719-Low-41250-High-25334-High-23876-[HiveSpec] Hive to create SyncSet resource[Serial]", func() {
testCaseID := "23040"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create SyncSet for resource apply......")
syncSetName := testCaseID + "-syncset1"
configMapName := testCaseID + "-configmap1"
configMapNamespace := testCaseID + "-" + getRandomString() + "-hive1"
resourceMode := "Sync"
syncTemp := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource := syncSetResource{
name: syncSetName,
namespace: oc.Namespace(),
namespace2: configMapNamespace,
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: resourceMode,
template: syncTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName})
syncResource.create(oc)
e2e.Logf("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check if syncSet is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName, ok, DefaultTimeout, []string{"SyncSet", syncSetName, "-n", oc.Namespace()}).check(oc)
exutil.By("Test Syncset Resource part......")
e2e.Logf("OCP-34719, step 3: Check if clustersync and clustersynclease are created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterSyncLease", cdName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("OCP-42113: Check if there is STATUS in clustersync tabular output.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "STATUS", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "MESSAGE", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o", "wide"}).check(oc)
e2e.Logf("OCP-34719, step 4: Check clustersync will record all syncsets first success time.")
successMessage := "All SyncSets and SelectorSyncSets have been applied to the cluster"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, successMessage, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Success", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].result}", syncSetName)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"}).check(oc)
e2e.Logf("OCP-34719, step 5: Check firstSuccessTime won't be changed when there are new syncset created.")
firstSuccessTime, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"))
o.Expect(err).NotTo(o.HaveOccurred())
syncSetName2 := testCaseID + "-syncset2"
configMapName2 := testCaseID + "-configmap2"
configMapNamespace2 := testCaseID + "-" + getRandomString() + "-hive2"
syncTemp2 := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource2 := syncSetResource{
name: syncSetName2,
namespace: oc.Namespace(),
namespace2: configMapNamespace2,
cdrefname: cdName,
ramode: resourceMode,
cmname: configMapName2,
cmnamespace: configMapNamespace2,
template: syncTemp2,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName2})
syncResource2.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName2, ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Success", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].result}", syncSetName2)}).check(oc)
updatedFirstSuccessTime, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"))
o.Expect(err).NotTo(o.HaveOccurred())
if !updatedFirstSuccessTime.Equal(firstSuccessTime) {
e2e.Failf("firstSuccessTime changed when new SyncSet is created")
}
e2e.Logf("Check if configMaps are stored in resourcesToDelete field in ClusterSync CR and they are applied on the target cluster.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapName, "-n", configMapNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapName2, "-n", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName2)}).check(oc)
e2e.Logf("OCP-34719, step 6: Check Resource can be deleted from target cluster via SyncSet when resourceApplyMode is Sync.")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"resourceApplyMode": "Sync"}}`}).check(oc)
patchYaml := `
spec:
resources:
- apiVersion: v1
kind: Namespace
metadata:
name: ` + configMapNamespace2
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
e2e.Logf("Check if ConfigMap %s has deleted from target cluster and clusterSync CR.", configMapName2)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", "-n", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName2)}).check(oc)
e2e.Logf("OCP-41250: Check Resource won't be deleted from target cluster via SyncSet when resourceApplyMode is Upsert.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Namespace", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"Namespace\")].name}", syncSetName2)}).check(oc)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"resourceApplyMode": "Upsert"}}`}).check(oc)
e2e.Logf("Check if resourcesToDelete field is gone in ClusterSync CR.")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete}", syncSetName2)}).check(oc)
e2e.Logf("Delete Namespace CR from SyncSet, check if Namespace is still exit in target cluster")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "json", "-p", `[{"op": "replace", "path": "/spec/resources", "value":[]}]`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"Namespace\")].name}", syncSetName2)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Namespace", configMapNamespace2}).check(oc)
e2e.Logf("OCP-34719, step 8: Create a bad SyncSet, check if there will be error message in ClusterSync CR.")
syncSetName3 := testCaseID + "-syncset3"
configMapName3 := testCaseID + "-configmap3"
configMapNamespace3 := testCaseID + "-" + getRandomString() + "-hive3"
syncTemp3 := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource3 := syncSetResource{
name: syncSetName3,
namespace: oc.Namespace(),
namespace2: configMapNamespace3,
cdrefname: cdName,
ramode: resourceMode,
cmname: configMapName3,
cmnamespace: "namespace-non-exist",
template: syncTemp3,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName3})
syncResource3.create(oc)
errorMessage := fmt.Sprintf("SyncSet %s is failing", syncSetName3)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName3, ok, DefaultTimeout, []string{"SyncSet", syncSetName3, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, errorMessage, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Failed")].message}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Failed")].status}`}).check(oc)
exutil.By("OCP-23876: Test Syncset Patch part......")
e2e.Logf("Create a test ConfigMap CR on target cluster.")
configMapNameInRemote := testCaseID + "-patch-test"
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "configmap", configMapNameInRemote, "--from-literal=foo=bar", "-n", configMapNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNameInRemote, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "bar", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
syncSetPatchName := testCaseID + "-syncset-patch"
syncPatchTemp := filepath.Join(testDataDir, "syncset-patch.yaml")
patchContent := `{ "data": { "foo": "baz-strategic" } }`
patchType := "strategic"
syncPatch := syncSetPatch{
name: syncSetPatchName,
namespace: oc.Namespace(),
cdrefname: cdName,
cmname: configMapNameInRemote,
cmnamespace: configMapNamespace,
pcontent: patchContent,
patchType: patchType,
template: syncPatchTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
syncPatch.create(oc)
e2e.Logf("Check if SyncSetPatch is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetPatchName, ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in strategic patch type.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "strategic", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-strategic", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in merge patch type.")
patchYaml = `
spec:
patches:
- apiVersion: v1
kind: ConfigMap
name: ` + configMapNameInRemote + `
namespace: ` + configMapNamespace + `
patch: |-
{ "data": { "foo": "baz-merge" } }
patchType: merge`
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "merge", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-merge", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in json patch type.")
patchYaml = `
spec:
patches:
- apiVersion: v1
kind: ConfigMap
name: ` + configMapNameInRemote + `
namespace: ` + configMapNamespace + `
patch: |-
[ { "op": "replace", "path": "/data/foo", "value": "baz-json" } ]
patchType: json`
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "json", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-json", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
exutil.By("OCP-25334: Test Syncset SecretReference part......")
syncSetSecretName := testCaseID + "-syncset-secret"
syncSecretTemp := filepath.Join(testDataDir, "syncset-secret.yaml")
sourceName := testCaseID + "-secret"
e2e.Logf("Create temp Secret in current namespace.")
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), sourceName})
err = oc.Run("create").Args("secret", "generic", sourceName, "--from-literal=testkey=testvalue", "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, ok, DefaultTimeout, []string{"Secret", sourceName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check Secret won't exit on target cluster before syncset-secret created.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Secret", "-n", configMapNamespace}).check(oc)
syncSecret := syncSetSecret{
name: syncSetSecretName,
namespace: oc.Namespace(),
cdrefname: cdName,
sname: sourceName,
snamespace: oc.Namespace(),
tname: sourceName,
tnamespace: configMapNamespace,
template: syncSecretTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetSecretName})
syncSecret.create(oc)
e2e.Logf("Check if syncset-secret is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetSecretName, ok, DefaultTimeout, []string{"SyncSet", syncSetSecretName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if the Secret is copied to the target cluster.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Secret", sourceName, "-n", configMapNamespace}).check(oc)
})
//author: [email protected]
//For simplicity, replace --simulate-bootstrap-failure with not copying aws-creds to make install failed
//example: ./bin/extended-platform-tests run all --dry-run|grep "62158"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-35990-[HiveSpec] Hive support limiting install attempt[Serial]", func() {
testCaseID := "35990"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
oc.SetupProject()
e2e.Logf("Don't copy AWS platform credentials to make install failed.")
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterDeployment with installAttemptsLimit=0...")
clusterTemp := filepath.Join(testDataDir, "clusterdeployment.yaml")
clusterLimit0 := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 0,
template: clusterTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterLimit0.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "InstallAttemptsLimitReached", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type==\"ProvisionStopped\")].reason}"}).check(oc)
o.Expect(checkResourceNumber(oc, cdName, []string{"pods", "-A"})).To(o.Equal(0))
exutil.By("Delete the ClusterDeployment and recreate it with installAttemptsLimit=1...")
cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterLimit1 := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: clusterTemp,
}
clusterLimit1.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "InstallAttemptsLimitReached", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type==\"ProvisionStopped\")].reason}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"pods", "-n", oc.Namespace()}).check(oc)
})
// Author: [email protected]
// ./bin/extended-platform-tests run all --dry-run|grep "63862"|./bin/extended-platform-tests run --timeout 80m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-63862-Medium-31931-[HiveSpec] MachinePool Supports Public Subnets[Serial]", func() {
// Describes a testing scenario
// azs: azs to put in the MachinePool's manifest
// subnets: subnets to put in the MachinePool's manifest
// expectedSubnets: subnets where we expect to find VM(s)
// expectedStatus: status of the InvalidSubnets condition
type scenario struct {
description string
azs []string
subnets []string
expectedSubnets []string
expectedStatus string
}
// Settings
var (
testCaseID = "63862"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
installConfigSecretName = cdName + "-install-config"
clusterImageSetName = cdName + "-imageset"
machinePoolNamePrefix = "infra"
machinePoolReplicas = 2
machinePoolCount = 0
stackName = "vpc-stack-" + testCaseID
azCount = 3
cidr = "10.0.0.0/16"
azsForTesting = []string{AWSRegion + "a", AWSRegion + "b"}
tmpDir = "/tmp/" + testCaseID + "-" + getRandomString()
)
// AWS Clients
var (
cfg = getAWSConfig(oc, AWSRegion)
cloudFormationClient = cloudformation.NewFromConfig(cfg)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
getSubnetByAz = func(subnetIds []string) map[string]string {
describeSubnetsOutput, err := ec2Client.DescribeSubnets(context.Background(), &ec2.DescribeSubnetsInput{
SubnetIds: subnetIds,
})
o.Expect(err).NotTo(o.HaveOccurred())
subnetByAz := make(map[string]string)
for _, subnet := range describeSubnetsOutput.Subnets {
az := aws.ToString(subnet.AvailabilityZone)
subnetId := aws.ToString(subnet.SubnetId)
if existingSubnet, ok := subnetByAz[az]; ok {
e2e.Failf("Subnet %v already exists in AZ %v !", existingSubnet, az)
}
subnetByAz[az] = subnetId
}
return subnetByAz
}
// Returns MachinePool.spec.name
getMachinePoolSpecName = func() string {
return machinePoolNamePrefix + "-" + strconv.Itoa(machinePoolCount)
}
// Returns MachinePool.metadata.name
getMachinePoolFullName = func() string {
return cdName + "-" + getMachinePoolSpecName()
}
createMachinePoolWithSubnetsAndAzs = func(subnets []string, azs []string) {
machinePoolName := getMachinePoolSpecName()
machinePoolManifest := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + "-" + machinePoolName + `
namespace: ` + oc.Namespace() + `
annotations:
# OCP-50051: Day 0 MachineSet Security Group Filters workaround
# We need this tag here as we will be creating VMs outside of the cluster's VPC
hive.openshift.io/extra-worker-security-group: default
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: ` + machinePoolName + `
replicas: ` + strconv.Itoa(machinePoolReplicas) + `
platform:
aws:
rootVolume:
size: 22
type: gp3
type: m5.xlarge`
if len(subnets) > 0 {
machinePoolManifest += "\n subnets:\n - " + strings.Join(subnets, "\n - ")
}
if len(azs) > 0 {
machinePoolManifest += "\n zones:\n - " + strings.Join(azs, "\n - ")
}
machinePoolYamlFile := tmpDir + "/" + machinePoolNamePrefix + ".yaml"
err := os.WriteFile(machinePoolYamlFile, []byte(machinePoolManifest), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
// No need to defer a deletion -- MachinePool VMs are deprovisioned along with the cluster
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", machinePoolYamlFile).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
}
getSubnetsForAzs = func(subnetsByAz map[string]string, azs []string) []string {
var subnets []string
for _, az := range azs {
switch subnet, ok := subnetsByAz[az]; ok {
case true:
subnets = append(subnets, subnet)
default:
e2e.Failf("For %v, no subnet found in AZ %v", subnetsByAz, az)
}
}
return subnets
}
checkMachinePoolStatus = func(target string) bool {
InvalidSubnetsCond :=
getCondition(oc, "MachinePool", getMachinePoolFullName(), oc.Namespace(), "InvalidSubnets")
if status, ok := InvalidSubnetsCond["status"]; !ok || status != target {
e2e.Logf("InvalidSubnets condition %v does not match target status %v, keep polling",
InvalidSubnetsCond, target)
return false
}
return true
}
checkMachinePoolVMsSubnets = func(kubeconfig string, expectedSubnets sets.Set[string]) bool {
// Make sure all instances are created
machinePoolInstancesIds := getMachinePoolInstancesIds(oc, getMachinePoolSpecName(), kubeconfig)
if len(machinePoolInstancesIds) != expectedSubnets.Len() {
e2e.Logf("%v MachinePool instances found (%v expected), keep polling",
len(machinePoolInstancesIds), len(expectedSubnets))
return false
}
// Make sure there's an instance in each expected subnet
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
InstanceIds: machinePoolInstancesIds,
})
o.Expect(err).NotTo(o.HaveOccurred())
for _, reservation := range describeInstancesOutput.Reservations {
instances := reservation.Instances
o.Expect(len(instances)).To(o.Equal(1))
instance := instances[0]
instanceId := aws.ToString(instance.InstanceId)
subnet := aws.ToString(instance.SubnetId)
o.Expect(expectedSubnets.Has(subnet)).To(o.BeTrue())
expectedSubnets.Delete(subnet)
e2e.Logf("Instance %v found in subnet %v", instanceId, subnet)
}
if expectedSubnets.Len() != 0 {
e2e.Logf("Expected subnets without VMs: %v", expectedSubnets.UnsortedList())
}
o.Expect(expectedSubnets.Len()).To(o.Equal(0))
return true
}
)
exutil.By("Creating Cluster")
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, clusterImageSetName, oc.Namespace(), installConfigSecretName, cdName)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Standing up a VPC and subnets which span multiple AZs")
endpointVpcTemp, err :=
testdata.Asset("test/extended/testdata/cluster_operator/hive/cloudformation-endpointvpc-temp.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
e2e.Logf("Deleting CloudFormation stack")
_, err := cloudFormationClient.DeleteStack(context.Background(), &cloudformation.DeleteStackInput{
StackName: aws.String(stackName),
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
e2e.Logf("Creating CloudFormation stack")
_, err = cloudFormationClient.CreateStack(context.Background(), &cloudformation.CreateStackInput{
StackName: aws.String(stackName),
TemplateBody: aws.String(string(endpointVpcTemp)),
Parameters: []cloudFormationTypes.Parameter{
{
ParameterKey: aws.String("AvailabilityZoneCount"),
ParameterValue: aws.String(strconv.Itoa(azCount)),
},
{
ParameterKey: aws.String("VpcCidr"),
ParameterValue: aws.String(cidr),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Making sure the CloudFormation stack is ready")
var vpcId, privateSubnetIds, publicSubnetIds string
waitUntilStackIsReady := func() bool {
describeStackOutput, err := cloudFormationClient.DescribeStacks(context.Background(),
&cloudformation.DescribeStacksInput{
StackName: aws.String(stackName),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeStackOutput.Stacks)).To(o.Equal(1))
stackStatus := describeStackOutput.Stacks[0].StackStatus
if stackStatus != cloudFormationTypes.StackStatusCreateComplete {
e2e.Logf("Stack status = %s, keep polling", stackStatus)
return false
}
// Get stack info once it is ready
for _, output := range describeStackOutput.Stacks[0].Outputs {
switch aws.ToString(output.OutputKey) {
case "VpcId":
vpcId = aws.ToString(output.OutputValue)
case "PrivateSubnetIds":
privateSubnetIds = aws.ToString(output.OutputValue)
case "PublicSubnetIds":
publicSubnetIds = aws.ToString(output.OutputValue)
}
}
return true
}
o.Eventually(waitUntilStackIsReady).WithTimeout(15 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
e2e.Logf("Found VpcId = %s, PrivateSubnetIds = %s, PublicSubnetIds = %s", vpcId, privateSubnetIds, publicSubnetIds)
e2e.Logf("Getting private/public subnets by AZ")
privateSubnetByAz := getSubnetByAz(strings.Split(privateSubnetIds, ","))
publicSubnetByAz := getSubnetByAz(strings.Split(publicSubnetIds, ","))
e2e.Logf("Public subnet by AZ = %v\nPrivate subnet by AZ = %v", publicSubnetByAz, privateSubnetByAz)
// We need to tag the default SG with key=Name since it does not come with a Name tag.
// This name will be used later in the hive.openshift.io/extra-worker-security-group: <sg-name>
// annotation on our MachinePool.
exutil.By("Tagging the default SG of the newly-created VPC")
describeSecurityGroupsOutput, err := ec2Client.DescribeSecurityGroups(context.Background(), &ec2.DescribeSecurityGroupsInput{
Filters: []types.Filter{
{
Name: aws.String("vpc-id"),
Values: []string{vpcId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
// According to our CloudFormation template, there should be one and only one SG -- the default SG
o.Expect(len(describeSecurityGroupsOutput.SecurityGroups)).To(o.Equal(1))
defaultSgId := aws.ToString(describeSecurityGroupsOutput.SecurityGroups[0].GroupId)
e2e.Logf("Found default SG = %v", defaultSgId)
_, err = ec2Client.CreateTags(context.Background(), &ec2.CreateTagsInput{
Resources: []string{defaultSgId},
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("default"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Default SG tagged")
exutil.By("Creating temporary directory")
defer func() {
_ = os.RemoveAll(tmpDir)
}()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Waiting for the CD to be installed")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Getting kubeconfig of the target cluster")
kubeconfig := getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
// For OCP-63862, we only test a number of allowed scenarios here.
exutil.By("Testing the following scenarios: ")
publicSubnetsForTesting := getSubnetsForAzs(publicSubnetByAz, azsForTesting)
privateSubnetsForTesting := getSubnetsForAzs(privateSubnetByAz, azsForTesting)
// Avoid modifying the first argument of append
blendedSubnetsForTesting := append([]string{publicSubnetsForTesting[0]}, privateSubnetsForTesting[1:]...)
allSubnetsForTesting := append([]string{}, publicSubnetsForTesting...)
allSubnetsForTesting = append(allSubnetsForTesting, privateSubnetsForTesting...)
scenarios := []scenario{
{
description: "MachinePool %v is created with some AZs & num-of-az public subnets. " +
"There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: publicSubnetsForTesting,
expectedSubnets: publicSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs & num-of-az private subnets. " +
"There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: privateSubnetsForTesting,
expectedSubnets: privateSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs & a combination of num-of-az " +
"public/private subnets. There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: blendedSubnetsForTesting,
expectedSubnets: blendedSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs and num-of-az public + num-of-az private subnets. " +
"There is a one-to-one relationship between the AZs and the public subnets. " +
"There is a one-to-one relationship between the AZs and the private subnets.",
azs: azsForTesting,
subnets: allSubnetsForTesting,
expectedSubnets: privateSubnetsForTesting,
expectedStatus: "False",
},
}
for _, scenario := range scenarios {
machinePoolCount++
e2e.Logf(scenario.description, getMachinePoolSpecName())
createMachinePoolWithSubnetsAndAzs(scenario.subnets, scenario.azs)
// Poll until the status of the InvalidSubnets condition is no longer unknown
newCheck("expect", "get", asAdmin, requireNS, compare, "Unknown", nok,
DefaultTimeout, []string{"MachinePool", getMachinePoolFullName(),
`-o=jsonpath={.status.conditions[?(@.type=="InvalidSubnets")].status}`}).check(oc)
// Make sure the status of the InvalidSubnets condition stays expected for a while
o.Consistently(checkMachinePoolStatus).
WithTimeout(2 * time.Minute).
WithPolling(30 * time.Second).
WithArguments(scenario.expectedStatus).
Should(o.BeTrue())
// Make sure VMs are created in expected subnets
o.Eventually(checkMachinePoolVMsSubnets).
WithTimeout(4*time.Minute).
WithPolling(30*time.Second).
WithArguments(kubeconfig, sets.New[string](scenario.expectedSubnets...)).
Should(o.BeTrue())
}
})
// Author: [email protected] [email protected]
// ./bin/extended-platform-tests run all --dry-run|grep "41212"|./bin/extended-platform-tests run --timeout 80m -f -
g.It("Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-41212-High-43751-Medium-57403-[HiveSDRosa] [AWSGov] Hive supports to install private cluster [Disruptive]", func() {
// Settings
var (
testCaseID = "41212"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
stackName = "endpointvpc-stack-" + testCaseID
// Should not overlap with the CIDR of the associate VPC
cidr = "10.1.0.0/16"
// Number of AZs for the endpoint VPC. Can be different from the equivalent for the associated VPC.
azCount = 3
dnsRecordType = "Alias"
)
// For OCP-43751
if isGovCloud {
dnsRecordType = "ARecord"
}
// AWS Clients
var (
cfg = getAWSConfig(oc, region)
cloudFormationClient = cloudformation.NewFromConfig(cfg)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
callCmd = func(cmd *exec.Cmd) error {
e2e.Logf("Calling command: %v", cmd)
out, err := cmd.CombinedOutput()
e2e.Logf("Command output: %s", out)
return err
}
waitForHiveadmissionRedeployment = func(initialHiveConfigGenInt int) bool {
// Make sure HiveConfig's generation is new
hiveConfigGen, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.metadata.generation}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
hiveConfigGenInt, err := strconv.Atoi(hiveConfigGen)
o.Expect(err).NotTo(o.HaveOccurred())
if hiveConfigGenInt <= initialHiveConfigGenInt {
e2e.Logf("HiveConfig generation (%v) <= initial HiveConfig generation (%v), keep polling",
hiveConfigGenInt, initialHiveConfigGenInt)
return false
}
// Make sure the generation is observed
hiveConfigGenObs, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.status.observedGeneration}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
hiveConfigGenObsInt, err := strconv.Atoi(hiveConfigGenObs)
o.Expect(err).NotTo(o.HaveOccurred())
if hiveConfigGenObsInt != hiveConfigGenInt {
e2e.Logf("HiveConfig observed generation (%v) != HiveConfig generation (%v), keep polling",
hiveConfigGenObsInt, hiveConfigGenInt)
return false
}
return true
}
checkCDConditions = func() bool {
awsPrivateLinkFailedCondition :=
getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "AWSPrivateLinkFailed")
if status, ok := awsPrivateLinkFailedCondition["status"]; !ok || status != "False" {
e2e.Logf("For condition AWSPrivateLinkFailed, status = %s, keep polling", status)
return false
}
awsPrivateLinkReadyCondition :=
getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "AWSPrivateLinkReady")
if status, ok := awsPrivateLinkReadyCondition["status"]; !ok || status != "True" {
e2e.Logf("For condition AWSPrivateLinkReady, status = %s, keep polling", status)
return false
}
return true
}
compareLocalAndRemoteAllowedPrincipals = func(vpceId string, localAllowedPrincipals []string) bool {
describeVpcEndpointServicePermissionsOutput, err := ec2Client.DescribeVpcEndpointServicePermissions(
context.Background(),
&ec2.DescribeVpcEndpointServicePermissionsInput{
ServiceId: aws.String(vpceId),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
var remoteAllowedPrincipals []string
for _, principal := range describeVpcEndpointServicePermissionsOutput.AllowedPrincipals {
remoteAllowedPrincipals = append(remoteAllowedPrincipals, *principal.Principal)
}
sort.Strings(remoteAllowedPrincipals)
sort.Strings(localAllowedPrincipals)
e2e.Logf("Local allowed principals = %v; remote allowed principals = %v",
localAllowedPrincipals, remoteAllowedPrincipals)
return reflect.DeepEqual(localAllowedPrincipals, remoteAllowedPrincipals)
}
)
exutil.By("Extracting Hiveutil")
tmpDir := "/tmp/" + testCaseID + "-" + getRandomString()
defer func(tempdir string) {
_ = os.RemoveAll(tempdir)
}(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
hiveutilPath := extractHiveutil(oc, tmpDir)
e2e.Logf("hiveutil extracted to %v", hiveutilPath)
exutil.By("Standing up an endpoint VPC and related resources")
endpointVpcTemp, err :=
testdata.Asset("test/extended/testdata/cluster_operator/hive/cloudformation-endpointvpc-temp.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
// No need to wait for stack deletion to finish. This will save us a couple of minutes.
e2e.Logf("Deleting CloudFormation stack")
_, err := cloudFormationClient.DeleteStack(context.Background(), &cloudformation.DeleteStackInput{
StackName: aws.String(stackName),
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
e2e.Logf("Creating CloudFormation stack")
_, err = cloudFormationClient.CreateStack(context.Background(), &cloudformation.CreateStackInput{
StackName: aws.String(stackName),
TemplateBody: aws.String(string(endpointVpcTemp)),
Parameters: []cloudFormationTypes.Parameter{
{
ParameterKey: aws.String("AvailabilityZoneCount"),
ParameterValue: aws.String(strconv.Itoa(azCount)),
},
{
ParameterKey: aws.String("VpcCidr"),
ParameterValue: aws.String(cidr),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Making sure the CloudFormation stack is ready")
var vpcId, privateSubnetIds string
waitUntilStackIsReady := func() bool {
describeStackOutput, err := cloudFormationClient.DescribeStacks(context.Background(),
&cloudformation.DescribeStacksInput{
StackName: aws.String(stackName),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeStackOutput.Stacks)).To(o.Equal(1))
stackStatus := describeStackOutput.Stacks[0].StackStatus
if stackStatus != cloudFormationTypes.StackStatusCreateComplete {
e2e.Logf("Stack status = %s, keep polling", stackStatus)
return false
}
// Get stack info once it is ready
for _, output := range describeStackOutput.Stacks[0].Outputs {
switch aws.ToString(output.OutputKey) {
case "VpcId":
vpcId = aws.ToString(output.OutputValue)
case "PrivateSubnetIds":
privateSubnetIds = aws.ToString(output.OutputValue)
}
}
return true
}
o.Eventually(waitUntilStackIsReady).WithTimeout(15 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
e2e.Logf("VpcId = %s, PrivateSubnetIds = %s", vpcId, privateSubnetIds)
// For OCP-43751:
// The hiveutil awsprivatelink add/remove commands filter out the private route tables through the Name tag.
// On AWS Gov cloud openshift clusters are installed into a BYO VPC,
// which is often created as part of a CloudFormation stack.
// There is no guarantee that the private route tables,
// which belong to the same CloudFormation stack,
// has a Name (or whatever pre-defined) tag.
// Consequently, we need to tag these route tables ourselves.
// This makes the test cases disruptive.
if isGovCloud {
infraId, err := exutil.GetInfraID(oc)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found infraID = %v", infraId)
// Get resources to tag
describeVpcsOutput, err := ec2Client.DescribeVpcs(context.Background(), &ec2.DescribeVpcsInput{
Filters: []types.Filter{
{
Name: aws.String("tag-key"),
Values: []string{fmt.Sprintf("kubernetes.io/cluster/%v", infraId)},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeVpcsOutput.Vpcs)).To(o.Equal(1))
associatedVpcId := aws.ToString(describeVpcsOutput.Vpcs[0].VpcId)
e2e.Logf("Found associated VPC ID = %v", associatedVpcId)
describeRouteTableOutput, err := ec2Client.DescribeRouteTables(context.Background(), &ec2.DescribeRouteTablesInput{
Filters: []types.Filter{
{
Name: aws.String("vpc-id"),
Values: []string{associatedVpcId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
associatedRouteTables := describeRouteTableOutput.RouteTables
o.Expect(len(associatedRouteTables)).NotTo(o.BeZero())
var privateAssociatedRouteTableIds []string
for _, associatedRouteTable := range associatedRouteTables {
associatedRouteTableId := aws.ToString(associatedRouteTable.RouteTableId)
e2e.Logf("Found associated route table %v", associatedRouteTableId)
for _, route := range associatedRouteTable.Routes {
if natGatewayId := aws.ToString(route.NatGatewayId); natGatewayId != "" {
e2e.Logf("Found a route targeting a NAT gateway, route table %v is private", associatedRouteTableId)
privateAssociatedRouteTableIds = append(privateAssociatedRouteTableIds, associatedRouteTableId)
break
}
}
}
o.Expect(len(privateAssociatedRouteTableIds)).NotTo(o.BeZero())
// Tagging
e2e.Logf("Tagging %v with Name = private", privateAssociatedRouteTableIds)
defer func() {
_, err := ec2Client.DeleteTags(context.Background(), &ec2.DeleteTagsInput{
Resources: privateAssociatedRouteTableIds,
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("private"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
_, err = ec2Client.CreateTags(context.Background(), &ec2.CreateTagsInput{
Resources: privateAssociatedRouteTableIds,
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("private"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
}
// Some (idempotent) awsprivatelink subcommands below are polled until succeed.
// Rationale:
// Calling an awsprivatelink subcommand immediately after another might fail
// due to etcd being only eventually consistent (as opposed to strongly consistent).
// In fact, awsprivatelink subcommands often starts off GETTING resources,
// which are processed and UPDATED before the command terminates.
// As a result, the later command might end up getting stale resources,
// causing the UPDATE request it makes to fail.
exutil.By("Setting up privatelink")
defer func() {
cmd := exec.Command(hiveutilPath, "awsprivatelink", "disable", "-d")
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
}()
// This is the first awsprivatelink subcommand, so no need to poll
cmd := exec.Command(
hiveutilPath, "awsprivatelink",
"enable",
"--creds-secret", "kube-system/aws-creds",
"--dns-record-type", dnsRecordType,
"-d")
err = callCmd(cmd)
o.Expect(err).NotTo(o.HaveOccurred())
// Sleep for a few seconds, so the first round of polling is more likely to succeed
time.Sleep(5 * time.Second)
// Get HiveConfig's generation, which will be used to make sure HiveConfig is updated.
initialHiveConfigGen, _, err := oc.AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.metadata.generation}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
initialHiveConfigGenInt, err := strconv.Atoi(initialHiveConfigGen)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found HiveConfig generation = %v after enabling awsprivatelink", initialHiveConfigGenInt)
e2e.Logf("Setting up endpoint VPC")
defer func() {
cmd := exec.Command(
hiveutilPath, "awsprivatelink",
"endpointvpc", "remove", vpcId,
"--creds-secret", "kube-system/aws-creds",
"-d",
)
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
}()
cmd = exec.Command(
hiveutilPath, "awsprivatelink",
"endpointvpc", "add", vpcId,
"--region", region,
"--creds-secret", "kube-system/aws-creds",
"--subnet-ids", privateSubnetIds,
"-d",
)
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
// It is necessary to wait for the re-deployment of Hive-admission, otherwise the CD gets rejected.
exutil.By("Waiting for the re-deployment of Hive-admission")
o.Eventually(waitForHiveadmissionRedeployment).
WithTimeout(3 * time.Minute).
WithPolling(1 * time.Minute).
WithArguments(initialHiveConfigGenInt).
Should(o.BeTrue())
// Wait until the new hiveadmission Deployment is available
err = oc.
AsAdmin().
WithoutNamespace().
Run("wait").
Args("deploy/hiveadmission", "-n", HiveNamespace, "--for", "condition=available=true", "--timeout=3m").
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterImageSet")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret")
// Reuse endpoint VPC here to save cloud resource.
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfigPrivateLink{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
publish: PublishInternal,
machineNetworkCidr: cidr,
privateSubnetId1: strings.Split(privateSubnetIds, ",")[0],
privateSubnetId2: strings.Split(privateSubnetIds, ",")[1],
privateSubnetId3: strings.Split(privateSubnetIds, ",")[2],
template: filepath.Join(testDataDir, "aws-install-config-privatelink.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Copying AWS credentials")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copying pull secret")
createPullSecret(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment")
clusterDeployment := clusterDeploymentPrivateLink{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
credRef: AWSCreds,
region: region,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment-aws-privatelink.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterDeployment.create(oc)
exutil.By("OCP-57403: Support to add AdditionalAllowedPrincipals for PrivateLink VPCE Services")
e2e.Logf("Waiting for awsprivatelink reconcile to be done")
err = oc.
AsAdmin().
Run("wait").
Args("ClusterDeployment", cdName, "--for", "condition=AWSPrivateLinkReady=true", "--timeout=20m").
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Getting default allowed principal")
defaultAllowedPrincipalJsonPath :=
"{.status.platformStatus.aws.privateLink.vpcEndpointService.defaultAllowedPrincipal}"
defaultAllowedPrincipal, _, err := oc.
AsAdmin().
Run("get").
Args("ClusterDeployment", cdName, "-o=jsonpath="+defaultAllowedPrincipalJsonPath).
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(defaultAllowedPrincipal).To(o.HavePrefix(fmt.Sprintf("arn:%v:iam::", awsPartition)))
e2e.Logf("Found defaultAllowedPrincipal = %s", defaultAllowedPrincipal)
e2e.Logf("Getting vpce ID")
vpceIdJsonPath :=
"{.status.platformStatus.aws.privateLink.vpcEndpointService.id}"
vpceId, _, err := oc.
AsAdmin().
Run("get").
Args("ClusterDeployment", cdName, "-o=jsonpath="+vpceIdJsonPath).
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(vpceId).To(o.HavePrefix("vpce-svc-"))
e2e.Logf("Found vpce ID = %s", vpceId)
e2e.Logf("Adding an additionalAllowedPrincipal")
additionalAllowedPrincipal := fmt.Sprintf("arn:%v:iam::301721915996:user/fakefxie", awsPartition)
additionalAllowedPrincipalPatch := `
spec:
platform:
aws:
privateLink:
additionalAllowedPrincipals:
- ` + additionalAllowedPrincipal
err = oc.
AsAdmin().
Run("patch").
Args("ClusterDeployment", cdName, "--type", "merge", "-p", additionalAllowedPrincipalPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Comparing local & remote allowed principals")
localAllowedPrincipals := []string{defaultAllowedPrincipal, additionalAllowedPrincipal}
o.Eventually(compareLocalAndRemoteAllowedPrincipals).
WithTimeout(5*time.Minute).
WithPolling(30*time.Second).
WithArguments(vpceId, localAllowedPrincipals).
Should(o.BeTrue())
e2e.Logf("Removing additionalAllowedPrincipals from CD")
err = oc.
AsAdmin().
Run("patch").
Args("ClusterDeployment", cdName, "--type=json", "-p",
`[{"op":"remove", "path": "/spec/platform/aws/privateLink/additionalAllowedPrincipals"}]`).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Comparing local & remote allowed principals again")
localAllowedPrincipals = []string{defaultAllowedPrincipal}
o.Eventually(compareLocalAndRemoteAllowedPrincipals).
WithTimeout(5*time.Minute).
WithPolling(30*time.Second).
WithArguments(vpceId, localAllowedPrincipals).
Should(o.BeTrue())
exutil.By("Back to OCP-41212: Waiting for installation to finish")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Checking CD.status")
o.Eventually(checkCDConditions).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
privateLinkStatus, _, err := oc.
AsAdmin().
Run("get").
Args("clusterdeployment", cdName, "-o", "jsonpath={.status.platformStatus.aws.privateLink}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Now CD.status.platformStatus.aws.privateLink looks like: \n%s", privateLinkStatus)
// Open question: should we check if the IDs in privateLinkStatus are correct ?
o.Expect(strings.Contains(privateLinkStatus, "hostedZoneID")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "vpcEndpointID")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "vpcEndpointService")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "defaultAllowedPrincipal")).To(o.BeTrue())
exutil.By("Making sure the private target cluster is not directly reachable")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
_, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "--kubeconfig", kubeconfig).Outputs()
o.Expect(err).To(o.HaveOccurred())
exutil.By("Making sure the target cluster is reachable from the Hive cluster")
// Due to the PrivateLink networking setup (through awsprivatelink subcommands called above),
// the target cluster can only be accessed from worker nodes of the Hive cluster.
// This is not a problem for the Hive operator, as its Pods are deployed on the worker nodes by default.
selectors := map[string]string{
"node-role.kubernetes.io/worker": "",
}
workerNodeNames := getNodeNames(oc, selectors)
kubeconfigByteSlice, err := os.ReadFile(kubeconfig)
o.Expect(err).NotTo(o.HaveOccurred())
// Ensure literal interpretation by Bash
kubeconfigSingleQuotedStr := "'" + string(kubeconfigByteSlice) + "'"
// Take care of the SCC setup
output, err := exutil.DebugNode(oc, workerNodeNames[0], "bash", "-c",
fmt.Sprintf("echo %s > kubeconfig; oc get co --kubeconfig kubeconfig", kubeconfigSingleQuotedStr))
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("cloud-credential"))
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "32223"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:liangli-Medium-32223-Medium-35193-High-23308-Hive ClusterDeployment Check installed and uninstalled [Serial]", func() {
testCaseID := "32223"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Get multi-arch image required for cross-architecture provisioning")
multiarchImage, err := exutil.GetLatestImage("multi", "ocp", "4-dev-preview-multi")
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
vmType: AWSVmTypeARM64,
arch: archARM64,
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, multiarchImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("test OCP-23308: Hive install log does not contain admin credentials, but contains REDACTED LINE OF OUTPUT")
provisionPodName := getProvisionPodNames(oc, cdName, oc.Namespace())[0]
cmd, stdout, err := oc.Run("logs").Args("-f", provisionPodName, "-c", "hive").BackgroundRC()
defer cmd.Process.Kill()
o.Expect(err).NotTo(o.HaveOccurred())
f := stdout.(*os.File)
defer f.Close()
targetLines := []string{
fmt.Sprintf("Access the OpenShift web-console here: https://console-openshift-console.apps.%v.%v\"", cdName, AWSBaseDomain),
"REDACTED LINE OF OUTPUT",
}
targetFound := assertLogs(f, targetLines, nil, 3*time.Minute)
o.Expect(targetFound).To(o.BeTrue())
exutil.By("test OCP-32223 check install")
provisionName, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.provisionRef.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(provisionName).NotTo(o.BeEmpty())
e2e.Logf("test OCP-32223 install")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "true", ok, DefaultTimeout, []string{"job", provisionName + "-provision", "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels.hive\\.openshift\\.io/install}"}).check(oc)
exutil.By("test OCP-35193 check uninstall")
e2e.Logf("get aws_access_key_id by secretName")
awsAccessKeyID, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "aws-creds", "-n", oc.Namespace(), "-o=jsonpath={.data.aws_access_key_id}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(provisionName).NotTo(o.BeEmpty())
e2e.Logf("Modify aws creds to invalid")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "-n", oc.Namespace(), "-p", `{"data":{"aws_access_key_id":null}}`, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("delete ClusterDeployment")
_, _, _, err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Background()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationFailed", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationFailed", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].reason}`}).check(oc)
e2e.Logf("Change aws creds to valid again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "-n", oc.Namespace(), "-p", `{"data":{"aws_access_key_id":"`+awsAccessKeyID+`"}}`, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationSucceeded", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationSucceeded", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].reason}`}).check(oc)
exutil.By("test OCP-32223 check uninstall")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "true", ok, DefaultTimeout, []string{"job", cdName + "-uninstall", "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels.hive\\.openshift\\.io/uninstall}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "33642"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-33642-[HiveSpec] Hive supports cluster hibernation [Serial]", func() {
testCaseID := "33642"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check AWS ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check CD has Hibernating condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
exutil.By("patch the CD to Hibernating...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Hibernating"}}`}).check(oc)
e2e.Logf("Wait for CD to be Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
exutil.By("patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "63275"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-63275-[HiveSpec] Hive support for AWS IMDSv2 [Serial]", func() {
var (
testCaseID = "63275"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
workerMpName = "worker"
infraMpName = "infra"
infraMpName2 = "infra-2"
)
exutil.By("Creating ClusterDeployment")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
clusterDeployment := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, clusterDeployment.name+"-imageset", oc.Namespace(), installConfigSecret.name1, clusterDeployment.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, clusterDeployment)
exutil.By("Wait for the cluster to be installed")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Creating temporary directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer func() {
_ = os.RemoveAll(tmpDir)
}()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Saving kubeconfig of the target cluster")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
exutil.By("Creating worker MachinePool with metadataService.authentication un-specified")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: filepath.Join(testDataDir, "machinepool-worker-aws.yaml"),
}
workermp.create(oc)
exutil.By("Creating infra MachinePool with metadataService.authentication = Optional")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
authentication: "Optional",
template: filepath.Join(testDataDir, "machinepool-infra-aws.yaml"),
}
defer cleanupObjects(oc, objectTableRef{
"MachinePool", oc.Namespace(), fmt.Sprintf("%s-%s", cdName, infraMpName),
})
inframp.create(oc)
exutil.By("Creating another infra MachinePool with metadataService.authentication = Required")
fullInframpName2 := fmt.Sprintf("%s-%s", cdName, infraMpName2)
inframp2 := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + fullInframpName2 + `
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: ` + infraMpName2 + `
platform:
aws:
metadataService:
authentication: Required
rootVolume:
size: 22
type: gp2
type: m4.xlarge
replicas: 1`
filename := tmpDir + "/" + testCaseID + infraMpName2
err = os.WriteFile(filename, []byte(inframp2), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), fullInframpName2})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Extracting Instance IDs")
instanceIdByMachinePool := make(map[string]string)
machinePools := []string{workerMpName, infraMpName, infraMpName2}
getInstanceIds := func() bool {
for _, machinePool := range machinePools {
instanceIds := getMachinePoolInstancesIds(oc, machinePool, kubeconfig)
if len(instanceIds) == 0 {
e2e.Logf("%s Machines not found, keep polling", machinePool)
return false
}
instanceIdByMachinePool[machinePool] = instanceIds[0]
}
return true
}
o.Eventually(getInstanceIds).WithTimeout(10 * time.Minute).WithPolling(30 * time.Second).Should(o.BeTrue())
e2e.Logf("Instance IDs found: %v", instanceIdByMachinePool)
exutil.By("Checking IMDSv2 settings")
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
expectedIMDSv2 := map[string]string{
workerMpName: "optional",
infraMpName: "optional",
infraMpName2: "required",
}
for machinePool, instanceId := range instanceIdByMachinePool {
e2e.Logf("Checking IDMSv2 settings on a %s instance", machinePool)
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
InstanceIds: []string{instanceId},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeInstancesOutput.Reservations)).To(o.Equal(1))
o.Expect(len(describeInstancesOutput.Reservations[0].Instances)).To(o.Equal(1))
o.Expect(string(describeInstancesOutput.Reservations[0].Instances[0].MetadataOptions.HttpTokens)).
To(o.Equal(expectedIMDSv2[machinePool]))
// Limit the frequency of API calls
time.Sleep(5 * time.Second)
}
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "49471"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-49471-High-76460-High-23677- Change EC2RootVolume: make IOPS optional [Serial]", func() {
testCaseID := "49471"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret with iops=1...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool with IOPS optional ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
iops: 2,
template: workermachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
workermp.create(oc)
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
iops: 1,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
exutil.By("OCP-76460:Creating infra2 MachinePool with machineLabels configuration ......")
var infraMpName2 = "infra2"
tmpDir76460 := "/tmp/" + cdName + "-" + getRandomString()
defer func() {
_ = os.RemoveAll(tmpDir76460)
}()
err := os.MkdirAll(tmpDir76460, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
fullInframpName2 := fmt.Sprintf("%s-%s", cdName, infraMpName2)
inframp2 := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + fullInframpName2 + `
namespace: ` + oc.Namespace() + `
spec:
machineLabels:
mihuang-test-label: mihaung-test-value
machine.openshift.io/cluster-api-cluster: test-conflict
clusterDeploymentRef:
name: ` + cdName + `
name: ` + infraMpName2 + `
platform:
aws:
rootVolume:
size: 120
type: gp3
type: m6a.xlarge
replicas: 1`
filename := tmpDir76460 + "/" + testCaseID + infraMpName2
err = os.WriteFile(filename, []byte(inframp2), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), fullInframpName2})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Check worker machinepool .spec.platform.aws.rootVolume.iops = 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.rootVolume.iops}"}).check(oc)
e2e.Logf("Check infra machinepool .spec.platform.aws.rootVolume.iops = 1")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.rootVolume.iops}"}).check(oc)
exutil.By("OCP-23677: Allow modification of machine pool labels and taints")
e2e.Logf("Patching machinepool ...")
patchYaml := `
spec:
taints:
- effect: foo
key: bar
labels:
baz: qux`
err = oc.AsAdmin().Run("patch").Args("MachinePool", cdName+"-worker", "--type", "merge", "-p", patchYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Extracting kubeconfig from remote cluster ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
infraID, _, err := oc.AsAdmin().Run("get").Args("cd", cdName, "-o", "jsonpath='{.spec.clusterMetadata.infraID}'").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
infraID = strings.Trim(infraID, "'")
machineSetName := infraID + "-worker-" + AWSRegion + "a"
e2e.Logf("Checking taints & labels on MachineSet %v ...", machineSetName)
expectedTaints := "{\"effect\":\"foo\",\"key\":\"bar\"}"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedTaints, ok, DefaultTimeout, []string{"MachineSet", machineSetName, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.spec.taints[0]}'"}).check(oc)
expectedLabels := "{\"baz\":\"qux\"}"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedLabels, ok, DefaultTimeout, []string{"MachineSet", machineSetName, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.spec.metadata.labels}'"}).check(oc)
exutil.By("OCP-76460: Check the machinepool's status.ownedMachineLabels, only labels that do not conflict with the installer should be marked, while conflicting labels should not be marked.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "mihuang-test-label", ok, DefaultTimeout, []string{"MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "machine.openshift.io/cluster-api-cluster", nok, DefaultTimeout, []string{"MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'"}).check(oc)
exutil.By("OCP-76460: Log in to the spoke cluster and check the machinesets. They should ignore the conflicting labels generated by the installer code and add the new machine labels.")
machineSetName2 := infraID + "-" + infraMpName2 + "-" + AWSRegion + "a"
expectedLabels = "\"mihuang-test-label\":\"mihaung-test-value\""
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedLabels, ok, DefaultTimeout, []string{"MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels}'"}).check(oc)
e2e.Logf("Check the machinesets labels should ignore conflicts generated by the installer code.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, infraID, ok, DefaultTimeout, []string{"MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels.machine\\.openshift\\.io/cluster-api-cluster}'"}).check(oc)
exutil.By("OCP-76460: Update the machinelabels in the machinepool and check that the machinesets ignore the conflicts generated by the installer code.")
e2e.Logf("Patching infra2 machinepool ...")
patchYaml2 := `
spec:
machineLabels:
mihuangtest-conflict-label2: mihuang-test-conflict2
machine.openshift.io/cluster-api-machine-role: test-conflict`
err = oc.AsAdmin().Run("patch").Args("MachinePool", fullInframpName2, "--type", "merge", "-p", patchYaml2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("OCP-76460: Check that the machinepool's status.ownedMachineLabels updates correctly.")
o.Eventually(func() bool {
out, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'").Output()
if err != nil {
return false
}
if strings.Contains(out, "mihuangtest-conflict-label2") && !strings.Contains(out, "machine.openshift.io/cluster-api-machine-role") {
return true
}
return false
}, DefaultTimeout, 60*time.Second).Should(o.BeTrue())
e2e.Logf("OCP-76460: Log in to the spoke cluster and check that the machinesets update correctly.")
o.Eventually(func() bool {
out, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig="+kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels}'").Output()
if err != nil {
return false
}
if strings.Contains(out, "\"mihuangtest-conflict-label2\":\"mihuang-test-conflict2\"") && strings.Contains(out, fmt.Sprintf("\"machine.openshift.io/cluster-api-machine-role\":\"%s\"", infraMpName2)) {
return true
}
return false
}, DefaultTimeout, 60*time.Second).Should(o.BeTrue())
})
//author: [email protected] [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "24088"|./bin/extended-platform-tests run --timeout 90m -f -
g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-24088-Medium-33045-[HiveSpec] Provisioning clusters on AWS with managed dns [Serial]", func() {
testCaseID := "24088"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create Route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-33045 - Prevent ClusterDeployment deletion until managed DNSZone is gone")
exutil.By("Delete route53-aws-creds in hive namespace")
err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", "route53-aws-creds", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Try to delete cd")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
exutil.By("Check the deprovision pod is completed")
DeprovisionPodName := getDeprovisionPodName(oc, cdName, oc.Namespace())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Completed", ok, ClusterUninstallTimeout, []string{"pod", DeprovisionPodName, "-n", oc.Namespace()}).check(oc)
exutil.By("Check the cd is not removed")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
exutil.By("Check the dnszone is not removed")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
exutil.By("Create route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Wait until dnszone controller next reconcile, verify dnszone and cd are removed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "51195"|./bin/extended-platform-tests run --timeout 35m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-51195-DNSNotReadyTimeout should be terminal[Disruptive]", func() {
testCaseID := "51195"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Remove Route53-aws-creds in hive namespace if exists to make DNSNotReady")
cleanupObjects(oc, objectTableRef{"secret", HiveNamespace, "route53-aws-creds"})
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check DNSNotReady, Provisioned and ProvisionStopped condiitons")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "DNS Zone not yet available", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].message}`}).check(oc)
e2e.Logf("Check PROVISIONSTATUS=ProvisionStopped ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ProvisionStopped", ok, ClusterResumeTimeout+DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type=='Provisioned')].reason}"}).check(oc)
e2e.Logf("check ProvisionStopped=true and DNSNotReady.reason=DNSNotReadyTimedOut ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "DNSNotReadyTimedOut", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="ProvisionStopped")].status}`}).check(oc)
exutil.By("Check DNSNotReadyTimeOut beacuse the default timeout is 10 min")
creationTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.creationTimestamp}"))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get cluster create timestamp,creationTimestampp is %v", creationTimestamp)
dnsNotReadyTimedOuTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].lastProbeTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get dnsnotready timestap, dnsNotReadyTimedOuTimestamp is %v", dnsNotReadyTimedOuTimestamp)
difference := dnsNotReadyTimedOuTimestamp.Sub(creationTimestamp)
e2e.Logf("default timeout is %v mins", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically(">=", 10))
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "23676" | ./bin/extended-platform-tests run --timeout 40m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-23676-Create cluster with master terminated by manipulation[Serial]", func() {
testCaseID := "23676"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Getting infraID from CD...")
var infraID string
var err error
getInfraIDFromCD := func() bool {
infraID, _, err = oc.AsAdmin().Run("get").Args("cd", cdName, "-o=jsonpath={.spec.clusterMetadata.infraID}").Outputs()
return err == nil && strings.HasPrefix(infraID, cdName)
}
o.Eventually(getInfraIDFromCD).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
e2e.Logf("Found infraID = %v", infraID)
// Get AWS client
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
exutil.By("Waiting until the master VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilMasterVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{infraID + "-master-0", infraID + "-master-1", infraID + "-master-2"},
},
},
MaxResults: aws.Int32(6),
})
return err == nil && len(describeInstancesOutput.Reservations) == 3
}
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the master VMs...")
var instancesToTerminate []string
for _, reservation := range describeInstancesOutput.Reservations {
instancesToTerminate = append(instancesToTerminate, *reservation.Instances[0].InstanceId)
}
_, err = ec2Client.TerminateInstances(context.Background(), &ec2.TerminateInstancesInput{
InstanceIds: instancesToTerminate,
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Terminating master VMs %v", instancesToTerminate)
// The stage at which provision fails is not guaranteed. Here we just make sure provision actually fails.
exutil.By("Waiting for the first provision Pod to fail...")
provisionPod1 := getProvisionPodNames(oc, cdName, oc.Namespace())[0]
newCheck("expect", "get", asAdmin, requireNS, compare, "Failed", ok, 1800, []string{"pod", provisionPod1, "-o=jsonpath={.status.phase}"}).check(oc)
exutil.By("Waiting for the second provision Pod to be created...")
waitForProvisionPod2 := func() bool {
provisionPodNames := getProvisionPodNames(oc, cdName, oc.Namespace())
if len(provisionPodNames) > 1 {
e2e.Logf("The second provision Pod is created. Current provision Pods = %v", provisionPodNames)
return true
}
return false
}
o.Eventually(waitForProvisionPod2).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "23970" | ./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-23970-The cluster name is limited by 63 characters[Serial]", func() {
testCaseID := "23970"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Creating ClusterImageSet ...")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret ...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Creating pull-secret ...")
createPullSecret(oc, oc.Namespace())
exutil.By("Copying AWS credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment with a 64-character-long cluster name ...")
clusterName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen] + "-" + "123456789012345678901234567890123456789012345"
clusterDeployment := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: clusterName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
parameters := []string{
"--ignore-unknown-parameters=true",
"-f", clusterDeployment.template,
"-p", "FAKE=" + clusterDeployment.fake,
"NAME=" + clusterDeployment.name,
"NAMESPACE=" + clusterDeployment.namespace,
"BASEDOMAIN=" + clusterDeployment.baseDomain,
"CLUSTERNAME=" + clusterDeployment.clusterName,
"MANAGEDNS=" + strconv.FormatBool(clusterDeployment.manageDNS),
"PLATFORMTYPE=" + clusterDeployment.platformType,
"CREDREF=" + clusterDeployment.credRef,
"REGION=" + clusterDeployment.region,
"IMAGESETREF=" + clusterDeployment.imageSetRef,
"INSTALLCONFIGSECRET=" + clusterDeployment.installConfigSecret,
"PULLSECRETREF=" + clusterDeployment.pullSecretRef,
"INSTALLATTEMPTSLIMIT=" + strconv.Itoa(clusterDeployment.installAttemptsLimit),
}
// Manually create CD to capture the output of oc apply -f cd_manifest_file
var cfgFileJSON string
defer func() {
if err := os.RemoveAll(cfgFileJSON); err != nil {
e2e.Logf("Error removing file %v: %v", cfgFileJSON, err.Error())
}
}()
cfgFileJSON, err := oc.AsAdmin().Run("process").Args(parameters...).OutputToFile(getRandomString() + "-hive-resource-cfg.json")
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("apply").Args("-f", cfgFileJSON).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.ContainSubstring("Invalid cluster name (.spec.clusterName): must be no more than 63 characters"))
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "22382" | ./bin/extended-platform-tests run --timeout 10m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-22382-ClusterDeployment.spec cannot be changed during an update[Serial]", func() {
testCaseID := "22382"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("OCP-22382: clusterdeployment.spec does not allow edit during an update")
e2e.Logf("Make sure a provision Pod is created in the project's namespace")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "-provision-", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace()}).check(oc)
e2e.Logf("Now attempt to modify clusterdeployment.spec")
output, err := oc.AsAdmin().Run("patch").Args("cd", cdName, "--type=merge", "-p", "{\"spec\":{\"baseDomain\": \"qe1.devcluster.openshift.com\"}}").Output()
o.Expect(err).To(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("Attempted to change ClusterDeployment.Spec which is immutable"))
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "42721"|./bin/extended-platform-tests run --timeout 70m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-22379-Medium-42721-[HiveSpec] Adopt clusters to Hive [Serial]", func() {
testCaseID := "42721"
resourceNameSuffix := testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
e2e.Logf("Create ClusterImageSet")
imageSetName := "clusterimageset-" + resourceNameSuffix
clusterImageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
clusterImageSet.create(oc)
e2e.Logf("Copy AWS root credentials & pull-secret to the temporary namespace")
createAWSCreds(oc, oc.Namespace())
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool, wait for it to be ready")
poolName := "clusterpool-" + resourceNameSuffix
clusterPool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 2,
maxConcurrent: 2,
hibernateAfter: "3h",
template: filepath.Join(testDataDir, "clusterpool.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
clusterPool.create(oc)
newCheck("expect", "get", asAdmin, requireNS, compare, "2", ok, ClusterInstallTimeout, []string{"ClusterPool", poolName, "-o=jsonpath={.status.ready}"}).check(oc)
e2e.Logf("Get CDs in the ClusterPool")
CDsInPool := strings.Split(strings.Trim(getCDlistfromPool(oc, poolName), "\n"), "\n")
o.Expect(len(CDsInPool)).To(o.Equal(2))
// We will use the 2 CDs as another Hive cluster and the cluster to adopt respectively
hiveCluster2, clusterToAdopt := CDsInPool[0], CDsInPool[1]
e2e.Logf("Get kubeconfig of Hive cluster 2 (%v) and the cluster to adopt (%v)", hiveCluster2, clusterToAdopt)
tmpDir2 := "/tmp/" + hiveCluster2 + "-" + getRandomString()
defer os.RemoveAll(tmpDir2)
err := os.MkdirAll(tmpDir2, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, hiveCluster2, hiveCluster2, tmpDir2)
kubeconfig2 := tmpDir2 + "/kubeconfig"
tmpDirToAdopt := "/tmp/" + clusterToAdopt + "-" + getRandomString()
defer os.RemoveAll(tmpDirToAdopt)
err = os.MkdirAll(tmpDirToAdopt, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, clusterToAdopt, clusterToAdopt, tmpDirToAdopt)
kubeconfigToAdopt := tmpDirToAdopt + "/kubeconfig"
e2e.Logf("Get infra ID and cluster ID of the cluster to adopt")
infraID, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.status.infrastructureName}", "--kubeconfig", kubeconfigToAdopt).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
clusterID, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterversion", "version", "-o=jsonpath={.spec.clusterID}", "--kubeconfig", kubeconfigToAdopt).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found infra ID = %v, cluster ID = %v for the cluster to adopt", infraID, clusterID)
e2e.Logf(`Set up Hive cluster 2 (%v):
1) Deploy Hive
2) Copy AWS root credentials to the default namespace
3) Copy the pull-secret to the default namespace
4) Create a Secret containing the admin kubeconfig of the cluster to adopt in the default namespace
`, hiveCluster2)
// No need to set up a new project on Hive cluster 2 as it will eventually be de-provisioned.
// We will simply use the default namespace for this cluster.
// Likewise, there is no need to clean up the resources created on Hive cluster 2.
hiveCluster2NS := "default"
origKubeconfig := oc.GetKubeconf()
origAdminKubeconfig := exutil.KubeConfigPath()
origNS := oc.Namespace()
// Defer an anonymous function so that ALL (chained) setters are executed after running the test case.
// The deferred function is executed before all defers above, which means that the oc client object
// is restored (i.e. points back to Hive cluster 1) before cleaning up resources on that cluster.
// This is what we want.
defer func(origKubeconfig, origAdminKubeconfig, origNS string) {
oc.SetKubeconf(origKubeconfig).SetAdminKubeconf(origAdminKubeconfig).SetNamespace(origNS)
}(origKubeconfig, origAdminKubeconfig, origNS)
// From this point on, the oc client object points to Hive cluster 2.
oc.SetKubeconf(kubeconfig2).SetAdminKubeconf(kubeconfig2).SetNamespace(hiveCluster2NS)
// The installHiveOperator() function deploys Hive as admin. To deploy Hive on another cluster (Hive cluster 2 here), we have 3 options:
// 1) Create a new oc client object:
// This is complicated as we cannot use the NewCLI() function, which incorporates calls to beforeEach() and afterEach()
// and those two are disallowed in g.It(). Moreover, most fields of the utils.CLI type are internal and lack setters.
// 2) Use the existing oc client object, point it to Hive cluster 2, and make sure to restore it at the end.
// This is our approach here.
// 3) Modify the existing code s.t. Hive is deployed as non-admin (as guest for ex.):
// This is again complicated as we would need to alter the existing code infrastructure to a large extent.
installHiveOperator(oc, &ns, &og, &sub, &hc, testDataDir)
createAWSCreds(oc, hiveCluster2NS)
createPullSecret(oc, hiveCluster2NS)
adminKubeconfigSecretName := "admin-kubeconfig-adopt"
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", adminKubeconfigSecretName, "-n", hiveCluster2NS, "--from-file", kubeconfigToAdopt).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By(fmt.Sprintf("Adopt cluster %v into cluster %v", clusterToAdopt, hiveCluster2))
adoptCDName := clusterToAdopt + "-adopt"
adoptCD := clusterDeploymentAdopt{
name: adoptCDName,
namespace: hiveCluster2NS,
baseDomain: AWSBaseDomain,
adminKubeconfigRef: adminKubeconfigSecretName,
clusterID: clusterID,
infraID: infraID,
clusterName: adoptCDName,
manageDNS: false,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
// OCP-22379: Hive will abandon deprovision for any cluster when preserveOnDelete is true
preserveOnDelete: true,
template: filepath.Join(testDataDir, "clusterdeployment-adopt.yaml"),
}
adoptCD.create(oc)
exutil.By("Make sure the adopted CD is running on Hive cluster 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, 600, []string{"ClusterDeployment", adoptCDName, "-n", hiveCluster2NS, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Make sure SyncSet works on Hive cluster 2")
syncSetName := "syncset-" + resourceNameSuffix
configMapName := "configmap-" + resourceNameSuffix
configMapNamespace := "namespace-" + resourceNameSuffix
syncSetResource := syncSetResource{
name: syncSetName,
namespace: hiveCluster2NS,
namespace2: configMapNamespace,
cdrefname: adoptCDName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: "Sync",
template: filepath.Join(testDataDir, "syncset-resource.yaml"),
}
syncSetResource.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"cm", configMapName, "-n", configMapNamespace, "--kubeconfig", kubeconfigToAdopt}).check(oc)
exutil.By("Delete the adopted CD on Hive cluster 2")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", adoptCDName, "-n", hiveCluster2NS).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Make sure the adopted CD is gone on Hive cluster 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, adoptCDName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", hiveCluster2NS}).check(oc)
e2e.Logf("Make sure the cloud resources persist (here we look for the EC2 instances)")
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag-key"),
Values: []string{"kubernetes.io/cluster/" + infraID},
},
},
MaxResults: aws.Int32(6),
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeInstancesOutput.Reservations)).To(o.Equal(6))
})
//author: [email protected] [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "22381"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-High-22381-Medium-34882-High-24693-Hive additional machinepool test [Serial]", func() {
testCaseID := "34882"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("OCP-24693: Support a global pull secret override")
e2e.Logf("Granting temp user permission to create secret in Hive's namespace ...")
// This is done so that the createPullSecret function can be used on Hive's namespace
err := oc.AsAdmin().WithoutNamespace().Run("adm", "policy").Args("add-role-to-user", "edit", oc.Username(), "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Creating global pull-secret ...")
defer oc.AsAdmin().Run("delete").Args("secret", "pull-secret", "-n", HiveNamespace).Execute()
createPullSecret(oc, HiveNamespace)
e2e.Logf("Patching Hiveconfig ...")
patch := `
spec:
globalPullSecretRef:
name: pull-secret`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "-n=hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/globalPullSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "-n=hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
workermp.create(oc)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-22381: machinepool.spec.plaform does not allow edit")
e2e.Logf("Patch worker machinepool .spec.platform")
patchYaml := `
spec:
name: worker
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp3
type: m4.2xlarge`
output, err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("MachinePool", cdName+"-worker", "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml).Output()
o.Expect(err).To(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("field is immutable"))
e2e.Logf("Check machines type is still m4.xlarge")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "m4.xlarge", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.type}"}).check(oc)
exutil.By("OCP-34882: Hive should be able to create additional machinepool after deleting all MachinePools")
e2e.Logf("Delete all machinepools")
cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
e2e.Logf("Check there are no machinepools existing")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "No resources found", ok, DefaultTimeout, []string{"MachinePool", "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check there are no machinesets in remote cluster")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "No resources found", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api"}).check(oc)
e2e.Logf("Create one more infra machinepool, check it can be created")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s can be created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machineset %s is up", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check machines is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "28867"|./bin/extended-platform-tests run --timeout 120m -f -
g.It("Author:lwan-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-28867-Medium-41776-[HiveSpec] Hive Machinepool test for autoscale [Serial]", func() {
testCaseID := "28867"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-28867: Hive supports an optional autoscaler settings instead of static replica count")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Patch static replicas to autoscaler")
autoScalingMax := "12"
autoScalingMin := "10"
removeConfig := "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig := fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is minimum value %s", autoScalingMin)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 3 3", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number is minReplicas %s when low workload", autoScalingMin)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 10 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't equal to minReplicas 10")
patchYaml := `
spec:
scaleDown:
enabled: true
delayAfterAdd: 10s
delayAfterDelete: 10s
delayAfterFailure: 10s
unneededTime: 10s`
e2e.Logf("Add busybox in remote cluster and check machines will scale up to maxReplicas %s", autoScalingMax)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ClusterAutoscaler", "default", "--type", "merge", "-p", patchYaml}).check(oc)
workloadYaml := filepath.Join(testDataDir, "workload.yaml")
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "busybox", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Deployment", "busybox", "-n", "default"}).check(oc)
e2e.Logf("Check replicas will scale up to maximum value %s", autoScalingMax)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 4 4", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number will scale up to maxReplicas %s", autoScalingMax)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 12 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't scale up to maxReplicas 12 after workload up")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "12", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Delete busybox in remote cluster and check machines will scale down to minReplicas %s", autoScalingMin)
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas will scale down to minimum value %s", autoScalingMin)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 3 3", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number will scale down to minReplicas %s", autoScalingMin)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 10 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't scale down to minReplicas 10 after workload down")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
removeConfig = "[{\"op\": \"remove\", \"path\": \"/spec/autoscaling\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
replicas := "3"
staticConfig := fmt.Sprintf("{\"spec\": {\"replicas\": %s}}", replicas)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", staticConfig}).check(oc)
exutil.By("OCP-41776: Allow minReplicas autoscaling of MachinePools to be 0")
e2e.Logf("Check hive allow set minReplicas=0 without zone setting")
autoScalingMax = "3"
autoScalingMin = "0"
removeConfig = "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig = fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check hive allow set minReplicas=0 within zone setting")
infra2MachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-infra2
namespace: ` + oc.Namespace() + `
spec:
autoscaling:
maxReplicas: 3
minReplicas: 0
clusterDeploymentRef:
name: ` + cdName + `
labels:
node-role.kubernetes.io: infra2
node-role.kubernetes.io/infra2: ""
name: infra2
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp3
type: m4.xlarge
zones:
- ` + AWSRegion + `a
- ` + AWSRegion + `b
- ` + AWSRegion + `c
taints:
- effect: NoSchedule
key: node-role.kubernetes.io/infra2`
var filename = testCaseID + "-machinepool-infra2.yaml"
err = ioutil.WriteFile(filename, []byte(infra2MachinepoolYaml), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("-f", filename, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 4*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
})
//author: [email protected]
//For simplicity, replace --simulate-bootstrap-failure with give an invalid root secret to make install failed
//example: ./bin/extended-platform-tests run all --dry-run|grep "23289"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-High-23289-Medium-39813-Test hive reports install restarts in CD and Metric[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
testCaseID := "23289"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
oc.SetupProject()
e2e.Logf("Create a invalid aws creds make install failed.")
e2e.Logf("Modify aws creds to invalid")
err := oc.Run("create").Args("secret", "generic", AWSCreds, "--from-literal=aws_access_key_id=test", "--from-literal=aws_secret_access_key=test", "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterDeployment with installAttemptsLimit=3...")
clusterTemp := filepath.Join(testDataDir, "clusterdeployment.yaml")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: clusterTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("OCP-23289: Check hive reports current number of install job retries in cluster deployment status...")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installRestarts}"}).check(oc)
o.Expect(checkResourceNumber(oc, cdName, []string{"pods", "-A"})).To(o.Equal(3))
exutil.By("OCP-39813: Check provision metric reporting number of install restarts...")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query := "hive_cluster_deployment_provision_underway_install_restarts"
checkResourcesMetricValue(oc, cdName, oc.Namespace(), "3", token, thanosQuerierURL, query)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "27559"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-27559-[HiveSpec] hive controllers can be disabled through a hiveconfig option [Disruptive]", func() {
e2e.Logf("Add \"maintenanceMode: true\" in hiveconfig.spec")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", "json", "-p", `[{"op":"remove", "path": "/spec/maintenanceMode"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"maintenanceMode": true}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check modifying is successful")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.maintenanceMode}"}).check(oc)
exutil.By("Check hive-clustersync and hive-controllers pods scale down, hive-operator and hiveadmission pods are not affected.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", nok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", nok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
e2e.Logf("Patch hiveconfig.spec.maintenanceMode to false")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type", "merge", "-p", `{"spec":{"maintenanceMode": false}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Verify the hive-controller and hive-clustersync pods scale up and appear")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
testCaseID := "27559"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "44477"|./bin/extended-platform-tests run --timeout 30m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-44477-Medium-44474-Medium-44476-[HiveSpec] Change fields of a steady pool, all unclaimed clusters will be recreated[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
testCaseID := "44477"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
imageSetName2 := poolName + "-imageset-2"
imageSet2 := clusterImageSet{
name: imageSetName2,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName2})
imageSet2.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet", "-A", "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName2, ok, DefaultTimeout, []string{"ClusterImageSet", "-A", "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and gcp-credentials to target namespace for the clusterdeployment
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config template Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := poolName + "-install-config-template"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: poolName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "10m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 2, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
jsonPath := "-o=jsonpath={\"reason:\"}{.status.conditions[?(@.type==\"AllClustersCurrent\")].reason}{\",status:\"}{.status.conditions[?(@.type==\"AllClustersCurrent\")].status}"
expectedResult := "reason:ClusterDeploymentsCurrent,status:True"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
field := []string{"imageSetRef", "userTags", "InstallConfigSecretTemplateRef"}
var (
caseID string
patchYaml string
jsonPathTemp string
expectedResultTemp string
)
for _, v := range field {
switch v {
case "imageSetRef":
caseID = "OCP-44476"
patchYaml = `{"spec":{"imageSetRef":{"name":"` + imageSetName2 + `"}}}`
jsonPathTemp = `-o=jsonpath={.items[?(@.spec.clusterPoolRef.poolName=="` + poolName + `")].spec.provisioning.imageSetRef.name}`
expectedResultTemp = imageSetName2 + " " + imageSetName2
case "userTags":
caseID = "OCP-44474"
patchYaml = `{"spec":{"platform":{"aws":{"userTags":{"cluster_desc":"` + poolName + `"}}}}}`
//jsonPathTemp = `-o=jsonpath={.items[?(@.spec.clusterPoolRef.poolName=="` + poolName + `")].spec.platform.aws.userTags.cluster_desc}`
//expectedResultTemp = poolName + " " + poolName
case "InstallConfigSecretTemplateRef":
caseID = "OCP-44477"
patchYaml = `{"spec":{"installConfigSecretTemplateRef":{"name":"` + installConfigSecretName + `"}}}`
default:
g.Fail("Given field" + v + " are not supported")
}
exutil.By(caseID + ": Change " + v + " field of a steady pool, all unclaimed clusters will be recreated")
e2e.Logf("oc patch ClusterPool field %s", v)
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterPool", poolName, "-n", oc.Namespace(), "-p", patchYaml, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
expectedResult = "reason:SomeClusterDeploymentsStale,status:False"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
expectedResult = "reason:ClusterDeploymentsCurrent,status:True"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
if v == "imageSetRef" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResultTemp, ok, DefaultTimeout, []string{"ClusterDeployment", "-A", jsonPathTemp}).check(oc)
}
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
}
exutil.By("Check Metrics for ClusterPool...")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query := "hive_clusterpool_stale_clusterdeployments_deleted"
e2e.Logf("Check metric %s Value equal to 6", query)
checkResourcesMetricValue(oc, poolName, oc.Namespace(), "6", token, thanosQuerierURL, query)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "27770"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-27770-Hive should set Condition when given ClusterImageSet or image doesn't exist[Serial]", func() {
testCaseID := "27770"
cdName1 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
cdName2 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config cd1 Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName1 + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName1,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment1...")
clusterImageSetName1 := cdName1 + "-imageset" + "-non-exist"
cluster1 := clusterDeployment{
fake: "false",
name: cdName1,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName1,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName1,
installConfigSecret: cdName1 + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster1.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster1.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster1)
exutil.By("Creating cd2 install-config Secret ...")
installConfigSecretName := cdName2 + "-install-config"
installConfigSecret = installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName2,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Creating cd2 ClusterImageSet with WrongReleaseImage...")
clusterImageSetName2 := cdName2 + "-imageset"
WrongReleaseImage := "registry.ci.openshift.org/ocp/release:4.13.0-0.nightly-2023-02-26-081527-non-exist"
imageSet := clusterImageSet{
name: clusterImageSetName2,
releaseImage: WrongReleaseImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName2})
imageSet.create(oc)
exutil.By("Creating cd2 with an incomplete pull-secret ...")
cluster2 := clusterDeployment{
fake: "false",
name: cdName2,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName2,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName2,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName2})
cluster2.create(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the message 'ClusterImageSet clusterImageSetName is not available'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, fmt.Sprintf("ClusterImageSet %s is not available", clusterImageSetName1), ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].message}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the reason 'ClusterImageSetNotFound'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterImageSetNotFound", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].reason}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the status 'False'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].status}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'ClusterImageSetNotFound', return no output")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ClusterImageSetNotFound\")]}'"}).check(oc)
exutil.By("Check pod pf cd2, return the status 'failed with Init:ImagePullBackOff'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Init:ImagePullBackOff", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace(), "--selector", "hive.openshift.io/imageset=true", "--selector", fmt.Sprintf("hive.openshift.io/cluster-deployment-name=%s", cdName2), "--no-headers"}).check(oc)
exutil.By("Check cd2 conditions with type 'installImagesNotResolved',return the reason 'JobToResolveImagesFailed'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "JobToResolveImagesFailed", ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"InstallImagesNotResolved\")].reason}'"}).check(oc)
exutil.By("Check cd2 conditions with type 'RequirementsMet',return the status 'True'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"InstallImagesNotResolved\")].status}'"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "28845"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-High-28845-Hive give a way to override the API URL of managed cluster[Serial]", func() {
testCaseID := "28845"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config cd Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check install status...")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("edit the cd CRs apiURLOverride field with a vaild apiURL")
ValidApiUrl := "https://api." + cdName + ".qe.devcluster.openshift.com:6443"
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("cd", cdName, "-n", oc.Namespace(), "--type=merge", "-p", fmt.Sprintf("{\"spec\":{\"controlPlaneConfig\":{\"apiURLOverride\": \"%s\"}}}", ValidApiUrl)).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).To(o.ContainSubstring("clusterdeployment.hive.openshift.io/" + cdName + " patched"))
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ActiveAPIURLOverride\")].status}'"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterReachable", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ActiveAPIURLOverride\")].reason}'"}).check(oc)
exutil.By("edit the cd CRs apiURLOverride field with an invaild apiURL")
InvalidApiUrl := "https://api." + cdName + "-non-exist.qe.devcluster.openshift.com:6443"
stdout, _, err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("cd", cdName, "-n", oc.Namespace(), "--type=merge", "-p", fmt.Sprintf("{\"spec\":{\"controlPlaneConfig\":{\"apiURLOverride\": \"%s\"}}}", InvalidApiUrl)).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).To(o.ContainSubstring("clusterdeployment.hive.openshift.io/" + cdName + " patched"))
waitForAPIWaitFailure := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "ActiveAPIURLOverride")
if status, ok := condition["status"]; !ok || status != "False" {
e2e.Logf("For condition ActiveAPIURLOverride, expected status is False, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "ErrorConnectingToCluster" {
e2e.Logf("For condition ActiveAPIURLOverride, expected reason is ErrorConnectingToCluster, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || !strings.Contains(message, "no such host") {
e2e.Logf("For condition ActiveAPIURLOverride, expected message is no such host, actual reason is %v, retrying ...", message)
return false
}
e2e.Logf("For condition ActiveAPIURLOverride, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForAPIWaitFailure).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("edit the cd CRs apiURLOverride field with a vaild apiURL again")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, fmt.Sprintf("clusterdeployment.hive.openshift.io/"+cdName+" patched"), ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", fmt.Sprintf("{\"spec\":{\"controlPlaneConfig\":{\"apiURLOverride\": \"%s\"}}}", ValidApiUrl)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ActiveAPIURLOverride\")].status}'"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterReachable", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ActiveAPIURLOverride\")].reason}'"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "32007"|./bin/extended-platform-tests run --timeout 20m -f -
g.It("Author:kcui-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-32007-Hive can prevent cluster deletion accidentally via a set on hiveconfig[Serial]", func() {
exutil.By("Add \"deleteProtection: enabled\" in hiveconfig.spec")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/deleteProtection"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"deleteProtection": "enabled"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check modifying is successful")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "enabled", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.deleteProtection}"}).check(oc)
testCaseID := "32007"
cdName1 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
cdName2 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config cd1 Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName1 + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName1,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment1...")
clusterImageSetName1 := cdName1 + "-imageset"
cluster1 := clusterDeployment{
fake: "true",
name: cdName1,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName1,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName1,
installConfigSecret: cdName1 + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster1.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster1.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster1)
exutil.By("Creating cd2 install-config Secret ...")
installConfigSecretName := cdName2 + "-install-config"
installConfigSecret = installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName2,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Creating cd2 ClusterImageSet")
clusterImageSetName2 := cdName2 + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName2,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName2})
imageSet.create(oc)
exutil.By("Creating cd2")
cluster2 := clusterDeployment{
fake: "true",
name: cdName2,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName2,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName2,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName2})
cluster2.create(oc)
exutil.By("Add annotations hive.openshift.io/protected-delete: \"false\" in cd2 CRs")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, fmt.Sprintf("clusterdeployment.hive.openshift.io/"+cdName2+" patched"), ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "--type", "merge", "-p", "{\"metadata\":{\"annotations\":{\"hive.openshift.io/protected-delete\": \"false\"}}}"}).check(oc)
exutil.By("Check Hive add the \"hive.openshift.io/protected-delete\" annotation to cd1 after installation")
e2e.Logf("Check cd1 is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, FakeClusterInstallTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.metadata.annotations.hive\\.openshift\\.io/protected-delete}'"}).check(oc)
exutil.By("delete cd1 will failed")
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName1, "-n", oc.Namespace()).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.ContainSubstring("metadata.annotations.hive.openshift.io/protected-delete: Invalid value: \"true\": cannot delete while annotation is present"))
exutil.By("edit hive.openshift.io/protected-delete: to \"false\" in cd1")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, fmt.Sprintf("clusterdeployment.hive.openshift.io/"+cdName1+" patched"), ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "--type", "merge", "-p", "{\"metadata\":{\"annotations\":{\"hive.openshift.io/protected-delete\": \"false\"}}}"}).check(oc)
exutil.By("delete cd1 again and success")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName1, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check cd1 has been deleted.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName1, nok, FakeClusterInstallTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
exutil.By("Check Hive didn't rewrite the \"hive.openshift.io/protected-delete\" annotation to cd2 after installation")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "false", ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "-o=jsonpath='{.metadata.annotations.hive\\.openshift\\.io/protected-delete}'"}).check(oc)
exutil.By("delete cd2 success")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName2, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check cd2 has been deleted.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName2, nok, FakeClusterInstallTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "29907"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-High-29907-Hive handles owner references after Velero restore[Serial]", func() {
testCaseID := "29907"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create Route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
e2e.Logf("Check dnszone has been created.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName+"-zone", ok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
exutil.By("check and record the messages of .metadata.ownerReferences1 and .metadata.resourceVersion1")
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("dnszone", cdName+"-zone", "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences[0]}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
var ownerReferences1 map[string]any
err = json.Unmarshal([]byte(stdout), &ownerReferences1)
o.Expect(err).NotTo(o.HaveOccurred())
resourceVersion1, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("dnszone", cdName+"-zone", "-n", oc.Namespace(), "-o=jsonpath={.metadata.resourceVersion}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("delete ownerReferences of the dnszone")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("dnszone", cdName+"-zone", "-n", oc.Namespace(), "--type=json", "-p", `[{"op":"remove", "path": "/metadata/ownerReferences"}]`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("check and record the messages of .metadata.ownerReferences2 and .metadata.resourceVersion2")
stdout, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("dnszone", cdName+"-zone", "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences[0]}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
var ownerReferences2 map[string]any
err = json.Unmarshal([]byte(stdout), &ownerReferences2)
o.Expect(err).NotTo(o.HaveOccurred())
resourceVersion2, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("dnszone", cdName+"-zone", "-n", oc.Namespace(), "-o=jsonpath={.metadata.resourceVersion}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("check the .metadata.ownerReferences is the same as before and the .metadata.resourceVersion is different")
CheckSameOrNot := func() bool {
if ownerReferences1["apiVersion"] == "" || ownerReferences1["blockOwnerDeletion"] != true || ownerReferences1["controller"] != true ||
ownerReferences1["kind"] != "ClusterDeployment" || ownerReferences1["name"] != cdName || ownerReferences1["uid"] == "" || resourceVersion1 == "" {
e2e.Logf("messages of ownerReferences1 or resourceVersion1 is wrong")
return false
}
if ownerReferences2["apiVersion"] == "" || ownerReferences2["blockOwnerDeletion"] != true || ownerReferences2["controller"] != true ||
ownerReferences2["kind"] != "ClusterDeployment" || ownerReferences2["name"] != cdName || ownerReferences2["uid"] == "" || resourceVersion2 == "" {
e2e.Logf("messages of ownerReferences2 or resourceVersion2 is wrong")
return false
}
if ownerReferences1["apiVersion"] != ownerReferences2["apiVersion"] || ownerReferences1["uid"] != ownerReferences2["uid"] || resourceVersion1 == resourceVersion2 {
e2e.Logf("ownerReferences1 or resourceVersion1 doesn't match the ownerReferences2 or resourceVersion2")
return false
}
return true
}
o.Eventually(CheckSameOrNot).WithTimeout(15 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "30089"|./bin/extended-platform-tests run --timeout 15m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-30089-Hive components will be teared down when HiveConfig is deleted[Disruptive]", func() {
exutil.By("Check the hive-controllers and hiveadmission are running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
exutil.By("Delete hiveconfig")
newCheck("expect", "delete", asAdmin, withoutNamespace, contain, "hiveconfig.hive.openshift.io \"hive\" deleted", ok, DefaultTimeout, []string{"hiveconfig", "hive"}).check(oc)
exutil.By("Check hive-controllers and hiveadmission were teared down or deleted")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", nok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", nok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", nok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", nok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", nok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", nok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
exutil.By("Create the hive resources again")
hc.createIfNotExist(oc)
exutil.By("Check the resources again")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pods", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"deployment", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"svc", "-n", "hive"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "35209"|./bin/extended-platform-tests run --timeout 45m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-35209-[HiveSpec] Allow setting lifetime for claims[Serial]", func() {
testCaseID := "35209"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 4,
maxSize: 4,
runningCount: 4,
maxConcurrent: 4,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
//the lifetime set for 4 claims initially
lifetimeMinuteInitials := []int{4, 8, 12, 20}
e2e.Logf("lifetimeMinuteInitials[] of four claims are %vm %vm(==default) %vm %vm(>maximum)", lifetimeMinuteInitials[0], lifetimeMinuteInitials[1], lifetimeMinuteInitials[2], lifetimeMinuteInitials[3])
defaultLifetimeMinute := 8
maximumLifetimeMinute := 16
e2e.Logf("defaultLifetimeMinute is %vm, maximumLifetimeMinute is %vm", defaultLifetimeMinute, maximumLifetimeMinute)
exutil.By("Add claimLifetime field (default and maximum) in .spec of clusterpool CR...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", fmt.Sprintf("{\"spec\":{\"claimLifetime\":{\"default\": \"%dm\"}}}", defaultLifetimeMinute)}).check(oc)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", fmt.Sprintf("{\"spec\":{\"claimLifetime\":{\"maximum\": \"%dm\"}}}", maximumLifetimeMinute)}).check(oc)
exutil.By("Check if ClusterPool has already existed")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
exutil.By("Create 4 clusterclaims named claim1 & claim2 & claim3 & claim4 with different .spec.lifetime from lifetimeMinuteInitials[]")
for claimIndex, lifetimeMinuteInitial := range lifetimeMinuteInitials {
exutil.By("Create a clusterclaim named claim" + strconv.Itoa(claimIndex+1))
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim" + strconv.Itoa(claimIndex+1)
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("patch claim" + strconv.Itoa(claimIndex+1) + " with spec.lifetime=" + strconv.Itoa(lifetimeMinuteInitial) + "m")
e2e.Logf("patch the lifetime if it not equals to defaultLifetimeMinute")
//if the .spec.lifetime is nil and default is not nil, it will be auto-filled by default lifetime
if lifetimeMinuteInitial != defaultLifetimeMinute {
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"clusterclaim", claimName, "-n", oc.Namespace(), "--type", "merge", "-p", fmt.Sprintf("{\"spec\":{\"lifetime\": \"%dm\"}}", lifetimeMinuteInitial)}).check(oc)
}
exutil.By("check the lifetime if it equals to lifetimeMinuteInitial[] or default or maximum lifetime")
//if the lifetimeMinuteSet > maximumLifetimeMinute, the liftime will be maximumLifetimeMinute, not the lifetimeMinuteSet
lifetimeMinuteFinal := int(math.Min(float64(lifetimeMinuteInitial), float64(maximumLifetimeMinute)))
newCheck("expect", "get", asAdmin, withoutNamespace, contain, fmt.Sprintf("%dm", lifetimeMinuteFinal), ok, DefaultTimeout, []string{"clusterclaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.status.lifetime}"}).check(oc)
}
//allowable for time error
timeThreshold := 30.0
//Check which claimName is timeout, between [0,4] is valid
timeoutClaimName := 0
//check each claimIndex status in different time
checkClaimStatus := func() bool {
//totally there are 4 claims, judge which claims should exist
if timeoutClaimName < 4 {
exutil.By(fmt.Sprintf("claim %d-4 should exist， check if it is really exist, by checking there are not deletionTimestamp", timeoutClaimName+1))
for claimNo := 4; claimNo > timeoutClaimName; claimNo-- {
claimName := poolName + "-claim" + strconv.Itoa(claimNo)
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterclaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.deletionTimestamp}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
//no deletionTimestamp means this claim still exist
o.Expect(stdout).To(o.Equal(""))
}
} else {
exutil.By("all claim should not exist, no need to check which claim still alive")
}
//there is no claim be end of life, return directly
if timeoutClaimName == 0 {
e2e.Logf("all claims exist, no need to check which claim disappears")
timeoutClaimName++
return true
}
//check the claim timeoutClaimName will be deleted in this time
exutil.By(fmt.Sprintf("check if claim 1-%d not exist or being deleted, only need to check the claim%v", timeoutClaimName, timeoutClaimName))
claimName := poolName + "-claim" + strconv.Itoa(timeoutClaimName)
//check if the claim has already been deleted
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterclaim", "-n", oc.Namespace()).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
//if the claim has been deleted, return directly
if !strings.Contains(stdout, claimName) {
e2e.Logf("the claim%d has been deleted, waiting for checking claim%d", timeoutClaimName, timeoutClaimName+1)
timeoutClaimName++
return true
}
//record creationTimestamp
stdout, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterclaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.creationTimestamp}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
creationTime, err := time.Parse(time.RFC3339, stdout)
o.Expect(err).NotTo(o.HaveOccurred())
//record deletionTimestamp
stdout, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterclaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.deletionTimestamp}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).NotTo(o.Equal(""))
deletionTime, err := time.Parse(time.RFC3339, stdout)
o.Expect(err).NotTo(o.HaveOccurred())
//calculate the lifetimeMinuteSet for this claimIndex
lifetimeMinuteFinal := int(math.Min(float64(lifetimeMinuteInitials[timeoutClaimName-1]), float64(maximumLifetimeMinute)))
//calculate the time error, and it should be less than the allowable time error set
gapTime := deletionTime.Sub(creationTime.Add(time.Duration(lifetimeMinuteFinal) * time.Minute))
o.Expect(math.Abs(gapTime.Seconds()) < timeThreshold).To(o.BeTrue())
timeoutClaimName++
return true
}
exutil.By("check the claim status on timeline")
o.Consistently(checkClaimStatus).WithTimeout(time.Duration(maximumLifetimeMinute+1) * time.Minute).WithPolling(time.Duration(lifetimeMinuteInitials[0]) * time.Minute).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "34148"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-34148-[HiveSpec] Hive supports spot instances in machine pools[Serial]", func() {
testCaseID := "34148"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config cd1 Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
clusterImageSetName := cdName + "-imageset"
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create spots instances")
replicasCount := 2
machinepoolName := cdName + "-spot"
spotMachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + machinepoolName + `
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: spot
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp2
type: m4.xlarge
spotMarketOptions: {}
replicas: ` + strconv.Itoa(replicasCount)
var filename = tmpDir + "/" + testCaseID + "-machinepool-spot.yaml"
err = os.WriteFile(filename, []byte(spotMachinepoolYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), machinepoolName})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Login to target cluster, check macheine & machineset are created on openshift-machine-api namespace.")
e2e.Logf("Extracting kubeconfig ...")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Checking the spotMachine number equals to replicas number: %v", replicasCount)
var instanceIds []string
checkSpotMachineCount := func() bool {
instanceIds = getMachinePoolInstancesIds(oc, "spot", kubeconfig)
e2e.Logf("spotMachineCount: %v", len(instanceIds))
return len(instanceIds) == replicasCount
}
o.Eventually(checkSpotMachineCount).WithTimeout(5 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
e2e.Logf("SpotMachine Instance IDs have been found")
e2e.Logf("Checking the spotMachineset ready number equals to replicas number: %v", replicasCount)
checkSpotMachinesetReadyCount := func() bool {
SpotMachinesetReadyCount := 0
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machineset", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
for _, spotMachinesetName := range strings.Split(stdout, " ") {
if strings.Contains(spotMachinesetName, "spot-") {
stdout, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machineset", spotMachinesetName, "-n", "openshift-machine-api", "-o=jsonpath={.status.replicas}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
tmpNumber, err := strconv.Atoi(stdout)
o.Expect(err).NotTo(o.HaveOccurred())
SpotMachinesetReadyCount += tmpNumber
}
}
e2e.Logf("spotMachinesetReadyCount: %v", SpotMachinesetReadyCount)
return SpotMachinesetReadyCount == replicasCount
}
o.Eventually(checkSpotMachinesetReadyCount).WithTimeout(5 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("login to aws console, check there will add 2 Spot Requests in ec2 \"Spot Requests\" list")
// Get AWS client
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
waitUntilSpotInstanceRequestsCreated := func() bool {
var describeSpotInstanceRequestsOutput *ec2.DescribeSpotInstanceRequestsOutput
describeSpotInstanceRequestsOutput, err = ec2Client.DescribeSpotInstanceRequests(context.Background(), &ec2.DescribeSpotInstanceRequestsInput{
Filters: []types.Filter{
{
Name: aws.String("instance-id"),
Values: instanceIds,
},
},
})
return err == nil && len(describeSpotInstanceRequestsOutput.SpotInstanceRequests) == 2
}
o.Eventually(waitUntilSpotInstanceRequestsCreated).WithTimeout(5 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Delete the machinepool")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("machinepool", machinepoolName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Checking the spotMachines disappear")
checkSpotMachineCount = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Count(stdout, "-spot-") == 0
}
o.Eventually(checkSpotMachineCount).WithTimeout(5 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Checking the spotMachineset ready number is 0")
checkSpotMachinesetReadyCount = func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machineset", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return strings.Count(stdout, "-spot-") == 0
}
o.Eventually(checkSpotMachinesetReadyCount).WithTimeout(5 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "62158"|./bin/extended-platform-tests run --timeout 30m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-High-62158-ClusterPool deletion should wait until all unclaimed clusters are destroyed - Case 1[Serial]", func() {
testCaseID := "62158"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
e2e.Logf("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
e2e.Logf("Check if ClusterPool created successfully and become ready")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Delete 1 CD, then one CD should be provisioning and another one CD should be provisioned")
cdNames := strings.Split(strings.Trim(getCDlistfromPool(oc, poolName), "\n"), "\n")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdNames[0], "-n", cdNames[0]).Background()
defer cmd.Process.Kill()
e2e.Logf("Checking the two CDs status...")
checkCDStatus := func() bool {
isProvisioned := 0
isProvisioning := 0
for _, cdName := range strings.Split(strings.Trim(getCDlistfromPool(oc, poolName), "\n"), "\n") {
stdout, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-n", cdName).Output()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(stdout, "Provisioned") {
isProvisioned++
}
if strings.Contains(stdout, "Provisioning") {
isProvisioning++
}
}
e2e.Logf("%v CD is Provisioned and %v CD is Provisioning", isProvisioned, isProvisioning)
return isProvisioned == 1 && isProvisioning == 1
}
o.Eventually(checkCDStatus).WithTimeout(300 * time.Second).WithPolling(5 * time.Second).Should(o.BeTrue())
exutil.By("Delete the ClusterPool, ClusterPool will be not deleted until both CDs are deleted")
stdout, err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterPool", poolName, "-n", oc.Namespace()).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).To(o.ContainSubstring("deleted"))
e2e.Logf("Check if all ClusterDeployments have been deleted")
o.Expect(getCDlistfromPool(oc, poolName)).To(o.Equal(""))
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "62158"|./bin/extended-platform-tests run --timeout 5m -f -
g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-69203-Add annotation to override installer image name. [Serial]", func() {
testCaseID := "69203"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
installerType := "installer-altinfra"
clusterImageSetName := cdName + "-imageset"
cluster := clusterDeployment{
fake: "false",
installerType: installerType,
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check ClusterDeployment installed pod is running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"pods", "-n", oc.Namespace(), "-l", "hive.openshift.io/job-type=provision", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
exutil.By("Check the image used is the version specified.")
secretFile, secretErr := getPullSecret(oc)
defer os.Remove(secretFile)
o.Expect(secretErr).NotTo(o.HaveOccurred())
installerImage, err := oc.AsAdmin().WithoutNamespace().Run("adm").Args("release", "info", "--image-for="+installerType, testOCPImage, "--registry-config="+secretFile).Output()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("testInstallerImage: %v", installerImage)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, installerImage, ok, DefaultTimeout, []string{"clusterdeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installerImage}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "62158"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-40825-[HiveSDRosa] Support AWS AssumeRole credentials cluster. [Disruptive]", func() {
testCaseID := "40825"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
hiveUserName := "hive_40825user"
hiveRoleName := "hive_40825role"
customerRoleName := "hive_40825csrole"
uuid := "abfgsheb765"
exutil.By("Prepare the AWS Assume Role needed for the test")
dirname := "/tmp/" + oc.Namespace() + "-" + testCaseID
err := os.MkdirAll(dirname, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(dirname)
e2e.Logf("Check if the user and roles exist")
cfg := getAWSConfig(oc, AWSRegion)
iamClient := iam.NewFromConfig(cfg)
_, err = iamClient.GetUser(context.Background(), &iam.GetUserInput{
UserName: aws.String(hiveUserName),
})
o.Expect(err).To(o.HaveOccurred())
roleNameList := []string{hiveRoleName, customerRoleName}
for _, roleName := range roleNameList {
_, err = iamClient.GetRole(context.Background(), &iam.GetRoleInput{
RoleName: aws.String(roleName),
})
o.Expect(err).To(o.HaveOccurred())
}
e2e.Logf("Create the user for hive testing")
defer func() {
_, err = iamClient.DeleteUser(context.Background(), &iam.DeleteUserInput{
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to delete the user")
}()
createUserOutput, err := iamClient.CreateUser(context.Background(), &iam.CreateUserInput{
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred())
hiveUserARN := createUserOutput.User.Arn
e2e.Logf("the user is successfully created: %v, the hiveUserArn is: %v", *createUserOutput.User.UserName, *hiveUserARN)
e2e.Logf("Assign the policy to the user")
_, err = iamClient.AttachUserPolicy(context.Background(), &iam.AttachUserPolicyInput{
PolicyArn: aws.String("arn:aws:iam::aws:policy/AdministratorAccess"),
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
attachedPolicyOutput, err := iamClient.ListAttachedUserPolicies(context.Background(), &iam.ListAttachedUserPoliciesInput{
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to list the attached policies")
for _, policy := range attachedPolicyOutput.AttachedPolicies {
_, err = iamClient.DetachUserPolicy(context.Background(), &iam.DetachUserPolicyInput{
PolicyArn: policy.PolicyArn,
UserName: aws.String(hiveUserName),
})
if err != nil {
e2e.Logf("failed to detach the policy: %v", err)
}
}
}()
time.Sleep(1 * time.Minute)
e2e.Logf("Create the role for hive user")
policyDocument, err := createAssumeRolePolicyDocument(*hiveUserARN, "")
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("the policyDocument is: %v", policyDocument)
defer func() {
_, err = iamClient.DeleteRole(context.Background(), &iam.DeleteRoleInput{
RoleName: aws.String(hiveRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to delete the hive role")
}()
createRoleOutput, err := iamClient.CreateRole(context.Background(), &iam.CreateRoleInput{
RoleName: aws.String(hiveRoleName),
AssumeRolePolicyDocument: aws.String(string(policyDocument)),
})
o.Expect(err).NotTo(o.HaveOccurred())
hiveRoleArn := createRoleOutput.Role.Arn
e2e.Logf("successfully created the role for hive testing: %v, hiveRoleArn is: %v", *createRoleOutput.Role.RoleName, *hiveRoleArn)
e2e.Logf(("Assign the policy to the role"))
_, err = iamClient.AttachRolePolicy(context.Background(), &iam.AttachRolePolicyInput{
PolicyArn: aws.String("arn:aws:iam::aws:policy/AdministratorAccess"),
RoleName: aws.String(hiveRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
attachedPolicies, err := iamClient.ListAttachedRolePolicies(context.Background(), &iam.ListAttachedRolePoliciesInput{
RoleName: aws.String(hiveRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to list the attached policies")
for _, policy := range attachedPolicies.AttachedPolicies {
_, err = iamClient.DetachRolePolicy(context.Background(), &iam.DetachRolePolicyInput{
PolicyArn: policy.PolicyArn,
RoleName: aws.String(hiveRoleName),
})
if err != nil {
e2e.Logf("failed to detach the policy: %v", err)
}
}
}()
time.Sleep(1 * time.Minute)
e2e.Logf("Create the customer role for hive role to assume")
customerRolePolicyDocument, err := createAssumeRolePolicyDocument(*hiveRoleArn, uuid)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("the customerRolePolicyDocument is: %v", customerRolePolicyDocument)
defer func() {
_, err := iamClient.DeleteRole(context.Background(), &iam.DeleteRoleInput{
RoleName: aws.String(customerRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to delete the customer role")
}()
createCustomerRoleOutput, err := iamClient.CreateRole(context.Background(), &iam.CreateRoleInput{
RoleName: aws.String(customerRoleName),
AssumeRolePolicyDocument: aws.String(string(customerRolePolicyDocument)),
})
o.Expect(err).NotTo(o.HaveOccurred())
customerRoleArn := createCustomerRoleOutput.Role.Arn
e2e.Logf("the created customer rolev %v for hive testing, role arn is: %v", *createCustomerRoleOutput.Role.RoleName, *customerRoleArn)
e2e.Logf("Attach the customer role to the hive role")
_, err = iamClient.AttachRolePolicy(context.Background(), &iam.AttachRolePolicyInput{
PolicyArn: aws.String("arn:aws:iam::aws:policy/AdministratorAccess"),
RoleName: aws.String(customerRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
attachedPolicies, err := iamClient.ListAttachedRolePolicies(context.Background(), &iam.ListAttachedRolePoliciesInput{
RoleName: aws.String(customerRoleName),
})
o.Expect(err).NotTo(o.HaveOccurred(), "failed to list the attached policies")
for _, policy := range attachedPolicies.AttachedPolicies {
_, err = iamClient.DetachRolePolicy(context.Background(), &iam.DetachRolePolicyInput{
PolicyArn: policy.PolicyArn,
RoleName: aws.String(customerRoleName),
})
if err != nil {
e2e.Logf("failed to detach the policy: %v", err)
}
}
}()
e2e.Logf("Create access key for hive user")
iamCredsOutput, err := iamClient.CreateAccessKey(context.Background(), &iam.CreateAccessKeyInput{
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred())
awsAccessKeyId := iamCredsOutput.AccessKey.AccessKeyId
awsSecretAccessKey := iamCredsOutput.AccessKey.SecretAccessKey
defer func() {
_, err = iamClient.DeleteAccessKey(context.Background(), &iam.DeleteAccessKeyInput{
AccessKeyId: aws.String(*awsAccessKeyId),
UserName: aws.String(hiveUserName),
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
e2e.Logf("Create aws-service-provider-config")
awsServiceProviderConfig := fmt.Sprintf(`
[default]
aws_access_key_id = %s
aws_secret_access_key = %s
role_arn = %s
`, *awsAccessKeyId, *awsSecretAccessKey, *hiveRoleArn)
awsServiceProviderConfigFile := filepath.Join(dirname, "aws-service-provider-config")
err = os.WriteFile(awsServiceProviderConfigFile, []byte(awsServiceProviderConfig), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Create aws-service-provider-secret to target namespace")
awsServiceProviderSecret := "aws-service-provider-secret"
defer oc.AsAdmin().Run("delete").Args("secret", awsServiceProviderSecret, "-n", HiveNamespace).Execute()
_, err = oc.AsAdmin().Run("create").Args("secret", "generic", awsServiceProviderSecret, "-n", HiveNamespace, "--from-file=aws_config="+awsServiceProviderConfigFile).Output()
o.Expect(err).NotTo(o.HaveOccurred())
secretOutput, err := oc.AsAdmin().Run("get").Args("secret", awsServiceProviderSecret, "-n", HiveNamespace).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(secretOutput).To(o.ContainSubstring(awsServiceProviderSecret))
e2e.Logf("Update HiveConfig to use the AWS Service Provider secret")
defer func() {
e2e.Logf("Restoring serviceProviderCredentialsConfig in HiveConfig")
restorePatch := `[{"op": "remove", "path": "/spec/serviceProviderCredentialsConfig"}]`
_, err := oc.AsAdmin().Run("patch").Args("hiveconfig", "hive", "--type", "json", "-p", restorePatch).Output()
o.Expect(err).NotTo(o.HaveOccurred())
}()
patchCmd := fmt.Sprintf("[{\"op\":\"replace\",\"path\":\"/spec/serviceProviderCredentialsConfig\",\"value\":{\"aws\":{\"credentialsSecretRef\":{\"name\":\"%s\"}}}}]", awsServiceProviderSecret)
_, err = oc.AsAdmin().Run("patch").Args(
"hiveconfig",
"hive",
"--type=json",
"-p="+patchCmd,
).Output()
o.Expect(err).NotTo(o.HaveOccurred())
secretRefOutput, err := oc.AsAdmin().Run("get").Args("hiveconfig", "hive", "-o=jsonpath={.spec.serviceProviderCredentialsConfig.aws.credentialsSecretRef.name}").Output()
e2e.Logf("secretRefOutput: %v", secretRefOutput)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(secretRefOutput).To(o.Equal("aws-service-provider-secret"))
exutil.By("Extract the ccoctl to create STS resources")
ccoctlTarget := "ccoctl"
ccoctlPath := exutil.ExtractCcoctl(oc, testOCPImage, ccoctlTarget)
defer os.Remove(filepath.Dir(ccoctlPath))
credsDir := filepath.Join(dirname, "creds")
e2e.Logf("Extract the credentials requests")
pullSecretFile, pullSecretErr := getPullSecret(oc)
o.Expect(pullSecretErr).NotTo(o.HaveOccurred())
defer os.Remove(pullSecretFile)
credsOutput, err := oc.AsAdmin().Run("adm").Args("release", "extract", testOCPImage, "--credentials-requests", "--cloud=aws", "--registry-config", pullSecretFile, "--to="+credsDir).Output()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("credsOutput: %v", credsOutput)
e2e.Logf("Create STS resources")
ccoctlOutputDir := filepath.Join(dirname, "_output")
defer func() {
e2e.Logf("Delete the STS resources")
deleteManifestsOutput, err := exec.Command("bash", "-c", fmt.Sprintf("%s aws delete --name %s --region %s", ccoctlPath, cdName, AWSRegion)).CombinedOutput()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("deleteManifestsOutput: %v", string(deleteManifestsOutput))
}()
createManifestsOutput, err := exec.Command("bash", "-c", fmt.Sprintf("%s aws create-all --name %s --region %s --credentials-requests-dir %s --output-dir %s", ccoctlPath, cdName, AWSRegion, credsDir, ccoctlOutputDir)).CombinedOutput()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("createManifestsOutput: %v", string(createManifestsOutput))
exutil.By("Create a Secret for your private service account signing key created with ccoctl aws create-all above.")
privateSAKeyName := "bound-service-account-signing-key"
defer func() {
e2e.Logf("Delete the Secret for your private service account signing key")
_, err := oc.AsAdmin().Run("delete").Args("secret", privateSAKeyName, "-n", oc.Namespace()).Output()
o.Expect(err).NotTo(o.HaveOccurred())
}()
_, err = oc.AsAdmin().Run("create").Args("secret", "generic", "-n", oc.Namespace(), privateSAKeyName, "--from-file=bound-service-account-signing-key.key="+filepath.Join(ccoctlOutputDir, "serviceaccount-signer.private")).Output()
o.Expect(err).NotTo(o.HaveOccurred())
boundServiceAccountSigningKeyOutput, err := oc.AsAdmin().Run("get").Args("secret", privateSAKeyName, "-n", oc.Namespace()).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(boundServiceAccountSigningKeyOutput).To(o.ContainSubstring(privateSAKeyName))
exutil.By("Create a Secret for installer manifests (credential role Secrets, Authentication config)")
manifestsSecretName := "cluster-manifests"
defer func() {
e2e.Logf("Delete the Secret for installer manifests")
_, err := oc.AsAdmin().Run("delete").Args("secret", manifestsSecretName, "-n", oc.Namespace()).Output()
o.Expect(err).NotTo(o.HaveOccurred())
}()
_, err = oc.AsAdmin().Run("create").Args("secret", "generic", manifestsSecretName, "-n", oc.Namespace(), "--from-file="+filepath.Join(ccoctlOutputDir, "manifests")).Output()
o.Expect(err).NotTo(o.HaveOccurred())
clusterManifestsOutput, err := oc.AsAdmin().Run("get").Args("secret", manifestsSecretName, "-n", oc.Namespace()).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(clusterManifestsOutput).To(o.ContainSubstring(manifestsSecretName))
exutil.By("Creating ClusterImageSet")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
credentialsMode: "Manual",
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Copying pull secret")
createPullSecret(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment")
clusterDeployment := clusterDeploymentAssumeRole{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
boundServiceAccountSigningKeySecretRef: privateSAKeyName,
clusterName: cdName,
platformType: "aws",
roleARN: *customerRoleArn,
externalID: uuid,
region: AWSRegion,
manifestsSecretRef: manifestsSecretName,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment-aws-assumerole.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterDeployment.create(oc)
exutil.By("Create worker MachinePool ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
)
workermp.create(oc)
exutil.By("Check ClusterDeployment installed pod is running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, 3*DefaultTimeout, []string{"pods", "-n", oc.Namespace(), "-l", "hive.openshift.io/job-type=provision", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check the worker machinepool replicas number equals to 3")
getClusterKubeconfig(oc, cdName, oc.Namespace(), dirname)
e2e.Logf("Check worker machinepool .status.replicas = 3")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
exutil.By("Patch machinepool static replicas to autoscaler")
autoScalingMax := "2"
autoScalingMin := "0"
removeConfig := "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig := fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Login to spoke cluster, check the MachineAutoscaler should be created")
kubeconfig := getClusterKubeconfig(oc, cdName, oc.Namespace(), dirname)
o.Eventually(func() bool {
machineAutoscalerNamesList, err := oc.AsAdmin().Run("get").Args("MachineAutoscaler", "-n", "openshift-machine-api", "-o", "jsonpath={.items[*].metadata.name}", "--kubeconfig="+kubeconfig).Output()
o.Expect(err).NotTo(o.HaveOccurred())
machineAutoscalerNames := strings.Fields(machineAutoscalerNamesList)
for _, machineAutoscalerName := range machineAutoscalerNames {
machineAutoscaler, _, err := oc.AsAdmin().Run("get").Args("MachineAutoscaler", machineAutoscalerName, "-n", "openshift-machine-api", "-o", "jsonpath={.spec.maxReplicas} { .spec.minReplicas}", "--kubeconfig="+kubeconfig).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
machineAutoscalerMax := "1"
machineAutoscalerMin := "0"
o.Expect(machineAutoscaler).To(o.Equal(machineAutoscalerMax + " " + machineAutoscalerMin))
}
return true
}).WithTimeout(2*time.Minute).WithPolling(10*time.Second).Should(o.BeTrue(), "MachineAutoscaler successfully created")
exutil.By("Patch machinepool autoscaler to static replicas")
removeConfig2 := "[{\"op\": \"remove\", \"path\": \"/spec/autoscaling\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig2}).check(oc)
recoverConfig := "{\"spec\": {\"replicas\": 2}}"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", recoverConfig}).check(oc)
e2e.Logf("Login to spoke cluster, check the MachineAutoscaler should be deleted")
o.Eventually(func() bool {
machineAutoscalerOutput, err := oc.AsAdmin().Run("get").Args("MachineAutoscaler", "-n", "openshift-machine-api", "-o", "jsonpath={.items[*].metadata.name}", "--kubeconfig="+kubeconfig).Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(machineAutoscalerOutput).To(o.BeEmpty())
return true
}).WithTimeout(2*time.Minute).WithPolling(10*time.Second).Should(o.BeTrue(), "MachineAutoscaler successfully deleted")
e2e.Logf("Check the machinepool replicas number equals to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run|grep "62158"|./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-78024-[HiveSpec] Support install cluster with ovn ipv4 subnet configured. [Serial]", func() {
testCaseID := "78024"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Config Install-Config Secret...")
ipv4InternalJoinSubnet := "101.64.0.0/16"
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
internalJoinSubnet: ipv4InternalJoinSubnet,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check AWS ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Log in to the spoke cluster, check that the OVN IPv4 internal subnet is correctly configured in the running cluster.")
e2e.Logf("Extracting kubeconfig ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
kubeconfigFilePath := getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
ovninternalJoinSubnet, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfigFilePath, "networks.operator.openshift.io", "cluster", "-o", "jsonpath={.spec.defaultNetwork.ovnKubernetesConfig.ipv4.internalJoinSubnet}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(ovninternalJoinSubnet).To(o.Equal(ipv4InternalJoinSubnet))
exutil.By("Verify whether the routing interface (LRP) of each node is using an IPv4 address.")
nodes, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfigFilePath, "nodes", "-o", "jsonpath={.items[*].metadata.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
nodeList := strings.Split(nodes, " ")
for _, node := range nodeList {
nodeGatewayRouterLrpIfaddrs, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfigFilePath, "node", node, "-o", "jsonpath={.metadata.annotations.k8s\\.ovn\\.org/node-gateway-router-lrp-ifaddrs}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(nodeGatewayRouterLrpIfaddrs).To(o.ContainSubstring("ipv4"))
re := regexp.MustCompile(`"ipv4":"([0-9\.\/]+)"`)
match := re.FindStringSubmatch(nodeGatewayRouterLrpIfaddrs)
o.Expect(len(match)).To(o.BeNumerically(">", 1))
ipWithCIDR := match[1]
ip := strings.Split(ipWithCIDR, "/")[0]
e2e.Logf("Node %s has gateway router LRP interface with IPv4 address %s", node, ip)
_, ipv4InternalJoinSubnetNet, err := net.ParseCIDR(ipv4InternalJoinSubnet)
o.Expect(err).NotTo(o.HaveOccurred())
nodeIP := net.ParseIP(ip)
o.Expect(nodeIP).NotTo(o.BeNil())
o.Expect(ipv4InternalJoinSubnetNet.Contains(nodeIP)).To(o.BeTrue(), fmt.Sprintf("Routing interface (LRP) of node %s has IPv4 address %s, but it is not in the expected subnet %s", node, ip, ipv4InternalJoinSubnet))
}
})
//author: [email protected]
//example: ./bin/extended-platform-tests run all --dry-run | grep "79046" | ./bin/extended-platform-tests run --timeout 60m -f -
g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-79046-[HiveSDRosa] AWS Non-CAPI CD install and Day2 infra MachinePool test. [Serial]", func() {
testCaseID := "79046"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
testImageVersion := "4.15"
testNonCAPIOCPImage, err := exutil.GetLatestNightlyImage(testImageVersion)
o.Expect(err).NotTo(o.HaveOccurred())
if testNonCAPIOCPImage == "" {
e2e.Failf("Failed to get image for version %v", testImageVersion)
}
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testNonCAPIOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create infra MachinePool ...")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Get the kubeconfig of the cluster")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check only 1 machineset up")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check only one machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 3")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 3}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale up to 3")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 3 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 2")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 2}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale down to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 2 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
})
})
| package hive | ||||
test case | openshift/openshift-tests-private | 34932084-668c-4fbf-a565-ec51423dce90 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-43974-Claims in excess of pool size should work well [Serial] | ['"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-43974-Claims in excess of pool size should work well [Serial]", func() {
// Settings
var (
testCaseID = "43974"
poolName = "pool-" + testCaseID
poolName2 = "pool2-" + testCaseID
imageSetName = poolName + "-imageset"
imageSetTemp = filepath.Join(testDataDir, "clusterimageset.yaml")
pool1claimName = poolName + "-claim"
pool1claimName2 = poolName + "-claim2"
pool1claimName3 = poolName + "-claim3"
pool2claimName = poolName2 + "-claim"
pool2claimName2 = poolName2 + "-claim2"
claimTemp = filepath.Join(testDataDir, "clusterclaim.yaml")
)
// Functions
var (
checkCDNumberinClusterPool = func(checkPoolName string, expectNum int) bool {
var cdNames []string
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("cd", "-A", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
for _, cdName := range strings.Split(stdout, " ") {
if strings.Contains(cdName, checkPoolName) {
cdNames = append(cdNames, cdName)
}
}
e2e.Logf("Number of cd in ClusterPool: %d, expectNum is %d", len(cdNames), expectNum)
return len(cdNames) == expectNum
}
checkCDinClusterPoolnotDeprovisioning = func(cdArray []string) bool {
for i := range cdArray {
cdName := cdArray[i]
condition := getCondition(oc, "cd", cdName, cdName, "Provisioned")
if reason, ok := condition["reason"]; !ok || reason == "Deprovisioning" {
e2e.Logf("For condition ProvisionFailed, expected reason can not be Deprovisioning, actual reason is %v, retrying ...", reason)
return false
}
}
return true
}
)
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 4,
runningCount: 0,
maxConcurrent: 4,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool"}).check(oc)
exutil.By("Check hive will create a cd")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName, 1).
Should(o.BeTrue())
exutil.By("Create claims more than pool size 1, here creating 3 clusterclaims from above pool")
pool1claim := clusterClaim{
name: pool1claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
pool1claim2 := clusterClaim{
name: pool1claimName2,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
pool1claim3 := clusterClaim{
name: pool1claimName3,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName2})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool1claimName3})
pool1claim.create(oc)
pool1claim2.create(oc)
pool1claim3.create(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName2, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool1claimName3, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
exutil.By("Check hive will create another 3 clusterdeployments for claims")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName, 4).
Should(o.BeTrue())
cdNameArray := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdNameArray), "\n")
exutil.By("Verify clusterdeployments won't be deleted before assigned to clusterclaims. Wait for cluster finishing installation and assignment, check they won't be deleted")
o.Consistently(checkCDinClusterPoolnotDeprovisioning).
WithTimeout(FakeClusterInstallTimeout * time.Second).
WithPolling(60 * time.Second).
WithArguments(cdArray).
Should(o.BeTrue())
exutil.By("Check all clusterclaims are assigned the cluster")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName2, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName, ok, DefaultTimeout, []string{"ClusterClaim", pool1claimName3, "-o=jsonpath={.spec.namespace}"}).check(oc)
exutil.By("Test when pool .spec.size=0, it works well too, create a clusterpool with .spec.size=0")
pool2 := clusterPool{
name: poolName2,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 0,
maxSize: 4,
runningCount: 0,
maxConcurrent: 4,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName2})
pool2.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterPool"}).check(oc)
exutil.By("check there will no cd created")
o.Consistently(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName2, 0).
Should(o.BeTrue())
exutil.By("Create 2 clusterclaim from above pool2")
pool2claim := clusterClaim{
name: pool2claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName2,
template: claimTemp,
}
pool2claim2 := clusterClaim{
name: pool2claimName2,
namespace: oc.Namespace(),
clusterPoolName: poolName2,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool2claimName})
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), pool2claimName2})
pool2claim.create(oc)
pool2claim2.create(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool2claimName, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, pool2claimName2, ok, DefaultTimeout, []string{"ClusterClaim"}).check(oc)
exutil.By("Check hive will create 2 clusterdeployments for claims")
o.Eventually(checkCDNumberinClusterPool).
WithTimeout(DefaultTimeout*time.Second).
WithPolling(5*time.Second).
WithArguments(poolName2, 2).
Should(o.BeTrue())
cdNameArray = getCDlistfromPool(oc, poolName2)
cdArray = strings.Split(strings.TrimSpace(cdNameArray), "\n")
exutil.By("Verify clusterdeployments won't be deleted before assigned to clusterclaims. Wait for cluster finishing installation and assignment, check they won't be deleted")
o.Consistently(checkCDinClusterPoolnotDeprovisioning).
WithTimeout(FakeClusterInstallTimeout * time.Second).
WithPolling(60 * time.Second).
WithArguments(cdArray).
Should(o.BeTrue())
exutil.By("Check all clusterclaims are assigned the cluster")
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterClaim", pool2claimName, "-o=jsonpath={.spec.namespace}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, contain, poolName2, ok, DefaultTimeout, []string{"ClusterClaim", pool2claimName2, "-o=jsonpath={.spec.namespace}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 8222e66f-28c7-45a5-9df6-fba32effffb7 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-46016-[HiveSpec] Test HiveConfig.Spec.FailedProvisionConfig.RetryReasons [Disruptive] | ['"context"', '"encoding/json"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-46016-[HiveSpec] Test HiveConfig.Spec.FailedProvisionConfig.RetryReasons [Disruptive]", func() {
// Settings
var (
testCaseID = "46016"
cdName1 = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
cdName2 = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
retryReasons = []string{"AWSVPCLimitExceeded", "S3BucketsLimitExceeded", "NoWorkerNodes", "UnknownError", "KubeAPIWaitFailed"}
)
// AWS Clients
var (
cfg = getAWSConfig(oc, region)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
TerminateVMs = func(describeInstancesOutput *ec2.DescribeInstancesOutput) error {
var instancesToTerminate []string
for _, reservation := range describeInstancesOutput.Reservations {
instancesToTerminate = append(instancesToTerminate, *reservation.Instances[0].InstanceId)
}
_, err := ec2Client.TerminateInstances(context.Background(), &ec2.TerminateInstancesInput{
InstanceIds: instancesToTerminate,
})
e2e.Logf("Terminating VMs %v", instancesToTerminate)
return err
}
)
exutil.By("Edit hiveconfig, add RetryReasons doesn't match cluster's failure")
patch := `
spec:
failedProvisionConfig:
retryReasons:
- AWSVPCLimitExceeded
- S3BucketsLimitExceeded`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/failedProvisionConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating Install-Config Secret...")
installConfigSecretName1 := cdName1 + "-install-config"
installConfigSecret := installConfig{
name1: cdName1 + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName1,
region: region,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName1,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName1,
platformType: "aws",
credRef: AWSCreds,
region: region,
imageSetRef: cdName1 + "-imageset",
installConfigSecret: installConfigSecretName1,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 2,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Getting infraID from CD...")
infraID := getInfraIDFromCDName(oc, cdName1)
exutil.By("Waiting until the Master & bootstrap VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilMasterVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{infraID + "-master-0", infraID + "-master-1", infraID + "-master-2", infraID + "-bootstrap"},
},
},
MaxResults: aws.Int32(8),
})
return err == nil && len(describeInstancesOutput.Reservations) == 4
}
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
// The stage at which provision fails is not guaranteed. Here we just make sure provision actually fails.
exutil.By("Waiting for the first provision Pod to fail...")
provisionPod1 := getProvisionPodNames(oc, cdName1, oc.Namespace())[0]
newCheck("expect", "get", asAdmin, requireNS, compare, "Failed", ok, ClusterInstallTimeout, []string{"pod", provisionPod1, "-o=jsonpath={.status.phase}"}).check(oc)
expectKeyValue := map[string]string{
"status": "True",
"reason": "FailureReasonNotRetryable",
}
waitForHRetryFailure1 := checkCondition(oc, "ClusterDeployment", cdName1, oc.Namespace(), "ProvisionStopped", expectKeyValue, "wait for cluster installment failure for FailureReasonNotRetryable")
o.Eventually(waitForHRetryFailure1).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
exutil.By("Edit hiveconfig, add RetryReasons match cluster's failure")
patch2 := `
spec:
failedProvisionConfig:
retryReasons:
- AWSVPCLimitExceeded
- S3BucketsLimitExceeded
- NoWorkerNodes
- UnknownError
- KubeAPIWaitFailed`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating Install-Config Secret...")
installConfigSecretName2 := cdName2 + "-install-config"
installConfigSecret2 := installConfig{
name1: cdName2 + "-install-config",
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName2,
region: region,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster2 := clusterDeployment{
fake: "false",
name: cdName2,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName2,
platformType: "aws",
credRef: AWSCreds,
region: region,
imageSetRef: cdName1 + "-imageset",
installConfigSecret: installConfigSecretName2,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 2,
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecret2.name1})
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName2})
installConfigSecret2.create(oc)
cluster2.create(oc)
exutil.By("Getting infraID from CD...")
infraID = getInfraIDFromCDName(oc, cdName2)
exutil.By("Waiting until the Master & bootstrap VMs are created...")
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Make sure the ProvisionFailed condition's reason matches one of retryReasons")
waitForProvisionFailed := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName2, oc.Namespace(), "ProvisionFailed")
if reason, ok := condition["reason"]; !ok || !ContainsInStringSlice(retryReasons, reason) {
e2e.Logf("For condition ProvisionFailed, expected reason is %v, actual reason is %v, retrying ...", retryReasons, reason)
return false
}
e2e.Logf("For condition ProvisionFailed, field reason matches one of retryReasons, proceeding to the next step ...")
return true
}
o.Eventually(waitForProvisionFailed).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
exutil.By("Getting infraID from CD again...")
var infraID2 string
getInfraIDFromCD3 := func() bool {
infraID2, _, err = oc.AsAdmin().Run("get").Args("cd", cdName2, "-o=jsonpath={.spec.clusterMetadata.infraID}").Outputs()
return err == nil && strings.HasPrefix(infraID2, cdName2) && infraID != infraID2
}
o.Eventually(getInfraIDFromCD3).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
infraID = infraID2
e2e.Logf("Found infraID = %v", infraID)
// Delete the machines again to make InstallAttempts reach the limit
exutil.By("Waiting until the Master & bootstrap VMs are created again ...")
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the Master & bootstrap VMs again ...")
err = TerminateVMs(describeInstancesOutput)
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue2 := map[string]string{
"status": "True",
"reason": "InstallAttemptsLimitReached",
}
waitForHRetryFailure2 := checkCondition(oc, "ClusterDeployment", cdName2, oc.Namespace(), "ProvisionStopped", expectKeyValue2, "wait for cluster installment failure for InstallAttemptsLimitReached")
o.Eventually(waitForHRetryFailure2).WithTimeout(20 * time.Minute).WithPolling(20 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | b588ac33-1327-4d49-915a-d967675a631f | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-59376-Configure resources on the hive deployment pods [Disruptive] | ['"encoding/json"', '"strings"', '"time"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-59376-Configure resources on the hive deployment pods [Disruptive]", func() {
exutil.By("Check the default spec.resources.requests.memory value of hive controller pod")
hiveControllersPod := getHivecontrollersPod(oc, HiveNamespace)
e2e.Logf("old hivecontrollers Pod is " + hiveControllersPod)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "512Mi", ok, DefaultTimeout, []string{"pods", hiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", hiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Edit hiveconfig, add deploymentConfig sections to HiveConfig's spec")
patch := `
spec:
deploymentConfig:
- deploymentName: hive-controllers
resources:
requests:
cpu: 50m
memory: 1024Mi`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/deploymentConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Hive controller pod will restart")
var newHiveControllersPod string
checkNewcontrollersPod := func() bool {
newHiveControllersPod = getHivecontrollersPod(oc, HiveNamespace)
return strings.Compare(hiveControllersPod, newHiveControllersPod) != 0
}
o.Eventually(checkNewcontrollersPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hivecontrollers Pod is " + newHiveControllersPod)
exutil.By("Check if the new deploymentConfig applied")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1Gi", ok, DefaultTimeout, []string{"pods", newHiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", newHiveControllersPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Configure deploymentConfig sections with empty resources.")
patch = `
spec:
deploymentConfig:
- deploymentName: hive-controllers
- deploymentName: hive-clustersync
- deploymentName: hiveadmission`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Verify hive-clustersync and hiveadmission pods, using the same method as hive-controllers")
hiveAdmissionPod := getHiveadmissionPod(oc, HiveNamespace)
e2e.Logf("old hiveadmission Pod is " + hiveAdmissionPod)
patch = `
spec:
deploymentConfig:
- deploymentName: hive-controllers
resources:
requests:
cpu: 50m
memory: 1024Mi
- deploymentName: hive-clustersync
resources:
requests:
cpu: 30m
memory: 600Mi
- deploymentName: hiveadmission
resources:
requests:
cpu: 50m
memory: 1024Mi`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
hiveClustersyncPod := "hive-clustersync-0"
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "600Mi", ok, DefaultTimeout, []string{"pods", hiveClustersyncPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "30m", ok, DefaultTimeout, []string{"pods", hiveClustersyncPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
var newHiveAdmissionPod string
checkNewadmissionPod := func() bool {
newHiveAdmissionPod = getHiveadmissionPod(oc, HiveNamespace)
return strings.Compare(hiveAdmissionPod, newHiveAdmissionPod) != 0
}
o.Eventually(checkNewadmissionPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hiveadmission Pod is " + newHiveAdmissionPod)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1Gi", ok, DefaultTimeout, []string{"pods", newHiveAdmissionPod, "-o=jsonpath={.spec.containers[0].resources.requests.memory}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "50m", ok, DefaultTimeout, []string{"pods", newHiveAdmissionPod, "-o=jsonpath={.spec.containers[0].resources.requests.cpu}", "-n", HiveNamespace}).check(oc)
exutil.By("Edit hiveconfig, add deploymentConfig sections to HiveConfig's spec with a bogus deploymentName, hiveconfig should not edit successfully and bounce immediately on schema validation")
patch = `
spec:
deploymentConfig:
- deploymentName: hive
resources:
requests:
cpu: 50m
memory: 1024Mi`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).To(o.HaveOccurred())
}) | |||||
test case | openshift/openshift-tests-private | 04d2fdf9-b87d-449b-8fcf-47e81d42658a | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41809-Formalize ClusterInstall Deletion Process [Disruptive] | ['"encoding/json"', '"os"', '"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41809-Formalize ClusterInstall Deletion Process [Disruptive]", func() {
testCaseID := "41809"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Configure hiveconfig to enable feature gate")
patch := `
spec:
featureGates:
custom:
enabled:
- AlphaAgentInstallStrategy
featureSet: Custom`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/featureGates"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create a fake cluster")
/*
make sure:
1. no need an install-config secret, remove it from file
2. no need .spec.provisioning field in ClusterDeployment CR, remove it
3. add .spec.clusterInstallRef field in ClusterDeployment CR
4. add a new FakeClusterInstall CR in the file
*/
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
fakeClusterYaml := `
apiVersion: v1
items:
- apiVersion: hive.openshift.io/v1
kind: ClusterDeployment
metadata:
name: ` + cdName + `
spec:
baseDomain: ` + AWSBaseDomain + `
clusterName: ` + cdName + `
controlPlaneConfig:
servingCertificates: {}
installed: false
platform:
aws:
credentialsSecretRef:
name: ` + AWSCreds + `
privateLink:
enabled: false
region: ` + AWSRegion + `
pullSecretRef:
name: ` + PullSecret + `
clusterInstallRef:
group: hiveinternal.openshift.io
version: v1alpha1
kind: FakeClusterInstall
name: ` + cdName + `
status: {}
- apiVersion: hiveinternal.openshift.io/v1alpha1
kind: FakeClusterInstall
metadata:
name: ` + cdName + `
spec:
imageSetRef:
name: ` + clusterImageSetName + `
clusterDeploymentRef:
name: ` + cdName + `
kind: List
metadata: {}`
var filename = tmpDir + "/" + testCaseID + "-fakecluster.yaml"
defer os.Remove(filename)
err = os.WriteFile(filename, []byte(fakeClusterYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"FakeClusterInstall", oc.Namespace(), cdName})
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, requireNS, contain, "true", ok, FakeClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check FakeClusterInstall CR, it has an ownerReferences related to the ClusterDeployment")
newCheck("expect", "get", asAdmin, requireNS, compare, cdName, ok, DefaultTimeout, []string{"FakeClusterInstall", cdName, "-o=jsonpath={.metadata.ownerReferences[0].name}"}).check(oc)
newCheck("expect", "get", asAdmin, requireNS, compare, "ClusterDeployment", ok, DefaultTimeout, []string{"FakeClusterInstall", cdName, "-o=jsonpath={.metadata.ownerReferences[0].kind}"}).check(oc)
exutil.By("Delete the clusterDeployment, check the FakeClusterInstall will be deleted too")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, requireNS, contain, cdName, nok, DefaultTimeout, []string{"FakeClusterInstall"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 0a4bade2-fe34-4085-8ae4-1310b8edfa60 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Low-39180-Hive MUST can modify statefulset spec.selector when given value is non-expected value even if this field is immutable [Disruptive] | ['"os"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Low-39180-Hive MUST can modify statefulset spec.selector when given value is non-expected value even if this field is immutable [Disruptive]", func() {
testCaseID := "39180"
resourceName := "test-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Scale down hive-operator")
// this will scale up the hive-operator and do some health check
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "{\"matchLabels\":{\"control-plane\":\"clustersync\",\"controller-tools.k8s.io\":\"1.0\"}}", ok, DefaultTimeout, []string{"sts", "-n", HiveNamespace, "-o=jsonpath={.items[].spec.selector}"}).check(oc)
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"pods", "hive-clustersync-0", "-n", HiveNamespace, "-o=jsonpath={.status.phase}"}).check(oc)
defer newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, DefaultTimeout, []string{"deployment/hive-operator", "-n", HiveNamespace, "-o=jsonpath={.status.readyReplicas}"}).check(oc)
defer oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployments", "hive-operator", "--replicas=1", "-n", HiveNamespace).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("scale").Args("deployments", "hive-operator", "--replicas=0", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Delete the existing statefulset CR")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("sts", "hive-clustersync", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Prepare a wrong statefulset CR")
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + resourceName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
statefulsetYaml := `
apiVersion: apps/v1
kind: StatefulSet
metadata:
labels:
app: hive
name: hive-clustersync
namespace: hive
spec:
replicas: 1
selector:
matchLabels:
app: hive
template:
metadata:
annotations:
labels:
app: hive
spec:
containers:
- image: quay.io/openshift-hive/hive:latest`
var filename = tmpDir + "/" + testCaseID + "-statefulset-hive.yaml"
err = os.WriteFile(filename, []byte(statefulsetYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename, "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"sts", "-n", HiveNamespace}).check(oc)
exutil.By("Check sts")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "{\"matchLabels\":{\"app\":\"hive\"}}", ok, DefaultTimeout, []string{"sts", "hive-clustersync", "-n", HiveNamespace, "-o=jsonpath={.spec.selector}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 438d50ee-875e-45a0-8999-8aa1cdf60746 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41525-Log diffs when validation rejects immutable modifications [Serial] | ['"fmt"', '"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-41525-Log diffs when validation rejects immutable modifications [Serial]", func() {
testCaseID := "41525"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Patch immutable fields of ClusterDeployment")
patchCDName := "test-cluster"
patchBaseDomain := "test.com"
patchRegion := "us-east-1"
patchimageSetRefName := "test-imageset"
patch := `
spec:
baseDomain: ` + patchBaseDomain + `
clusterName: ` + patchCDName + `
platform:
aws:
region: ` + patchRegion + `
provisioning:
imageSetRef:
name: ` + patchimageSetRefName
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, `--type=merge`, "-p", patch, "-n", oc.Namespace()).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.ContainSubstring("Attempted to change ClusterDeployment.Spec which is immutable"))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("ClusterName: (%s => %s)", cdName, patchCDName)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("BaseDomain: (%s => %s)", AWSBaseDomain, patchBaseDomain)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("Platform.AWS.Region: (%s => %s)", AWSRegion, patchRegion)))
o.Expect(stderr).To(o.ContainSubstring(fmt.Sprintf("Provisioning.ImageSetRef.Name: (%s => %s)", cdName+"-imageset", patchimageSetRefName)))
exutil.By("Check .spec of ClusterDeployment, the fields tried to be changed above didn't change,")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterName}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, AWSBaseDomain, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.baseDomain}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, AWSRegion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.region}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName+"-imageset", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.provisioning.imageSetRef}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 2a2361ab-dcac-47d1-ab80-b6fb12e2220d | Author:sguo-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-37464- Seperate clustersync controller from hive-controllers, meanwhile make it be able to scale up/down [Serial] | ['"encoding/json"', '"time"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:sguo-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-37464- Seperate clustersync controller from hive-controllers, meanwhile make it be able to scale up/down [Serial]", func() {
exutil.By("Check the statefulset in hive namespace")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"statefulset", "-n", HiveNamespace}).check(oc)
exutil.By("check there is a separate pod for clustersync")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-0", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Patching HiveConfig to scale up clustersync pod")
patch := `
spec:
controllersConfig:
controllers:
- config:
replicas: 2
name: clustersync`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/controllersConfig"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check statefulset replicas scale up to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "2", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Wait for 10 min to hive next reconcile finish, then check the hive-clustersync-1 pod is still there")
time.Sleep(10 * time.Minute)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "2", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
exutil.By("Scale down replicas to 1 again via editing hiveconfig, check it can scale down")
patch = `
spec:
controllersConfig:
controllers:
- config:
replicas: 1
name: clustersync`
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check statefulset replicas scale down to 1 again,")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, DefaultTimeout, []string{"statefulset", "hive-clustersync", "-o=jsonpath={.status.replicas}", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-0", ok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync-1", nok, DefaultTimeout, []string{"pods", "-n", HiveNamespace}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 060a4397-f882-493a-8753-a9b90ac4558a | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-43100-Hive supports hibernating AWS cluster with spot instances [Serial] | ['"context"', '"io/ioutil"', '"os"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-High-43100-Hive supports hibernating AWS cluster with spot instances [Serial]", func() {
testCaseID := "43100"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Create tmp directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Create spots instances, one with On-Demand and another with setting maxPrice")
spotMachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-spot
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: spot
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp2
type: m4.xlarge
spotMarketOptions: {}
replicas: 1`
var filename = tmpDir + "/" + testCaseID + "-machinepool-spot.yaml"
defer os.Remove(filename)
err = ioutil.WriteFile(filename, []byte(spotMachinepoolYaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-spot"})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
spotMachinepool2Yaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-spot2
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: spot2
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp2
type: m4.xlarge
spotMarketOptions:
maxPrice: "0.1"
replicas: 1`
var filename2 = tmpDir + "/" + testCaseID + "-machinepool-spot2.yaml"
defer os.Remove(filename2)
err = ioutil.WriteFile(filename2, []byte(spotMachinepool2Yaml), 0644)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-spot2"})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Login to target cluster, check spot instances are created")
e2e.Logf("Extracting kubeconfig ...")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
var oldSpotMachineName, oldSpotMachineName2 string
checkSpotMachineName := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("machine list: %s", stdout)
oldSpotMachineName = ""
oldSpotMachineName2 = ""
for _, MachineName := range strings.Split(stdout, " ") {
if strings.Contains(MachineName, "spot-") {
oldSpotMachineName = MachineName
}
if strings.Contains(MachineName, "spot2-") {
oldSpotMachineName2 = MachineName
}
}
e2e.Logf("oldSpotMachineName: %s, oldSpotMachineName2: %s", oldSpotMachineName, oldSpotMachineName2)
return strings.Contains(oldSpotMachineName, "spot-") && strings.Contains(oldSpotMachineName2, "spot2-")
}
o.Eventually(checkSpotMachineName).WithTimeout(DefaultTimeout * time.Second).WithPolling(5 * time.Second).Should(o.BeTrue())
// Get AWS client
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
e2e.Logf("Waiting until the spot VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilSpotVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{oldSpotMachineName, oldSpotMachineName2},
},
},
MaxResults: aws.Int32(6),
})
if err != nil {
e2e.Logf("Error when get describeInstancesOutput: %s", err.Error())
return false
}
e2e.Logf("Check result length: %d", len(describeInstancesOutput.Reservations))
for _, reservation := range describeInstancesOutput.Reservations {
instanceLen := len(reservation.Instances)
if instanceLen != 1 {
e2e.Logf("instanceLen should be 1, actual number is %d", instanceLen)
return false
}
e2e.Logf("Instance ID: %s, status: %s", *reservation.Instances[0].InstanceId, reservation.Instances[0].State.Name)
if reservation.Instances[0].State.Name != "running" {
e2e.Logf("Instances state should be running, actual state is %s", reservation.Instances[0].State.Name)
return false
}
}
return len(describeInstancesOutput.Reservations) == 2
}
o.Eventually(waitUntilSpotVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Hibernating the cluster and check ClusterDeployment Hibernating condition")
// the MachinePool can not be deleted when the ClusterDeployment is in Hibernating state
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue := map[string]string{
"status": "True",
"reason": "Hibernating",
"message": "Cluster is stopped",
}
waitForHibernating := checkCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "Hibernating", expectKeyValue, "wait for cluster hibernating")
o.Eventually(waitForHibernating).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
exutil.By("Check spot instances are terminated")
waitUntilSpotVMTerminated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{oldSpotMachineName, oldSpotMachineName2},
},
},
MaxResults: aws.Int32(6),
})
if err != nil {
e2e.Logf("Error when get describeInstancesOutput: %s", err.Error())
return false
}
e2e.Logf("Check result length: %d", len(describeInstancesOutput.Reservations))
for _, reservation := range describeInstancesOutput.Reservations {
instanceLen := len(reservation.Instances)
if instanceLen != 1 {
e2e.Logf("instanceLen should be 1, actual number is %d", instanceLen)
return false
}
e2e.Logf("Instance ID: %s, status: %s", *reservation.Instances[0].InstanceId, reservation.Instances[0].State.Name)
if reservation.Instances[0].State.Name != "terminated" {
e2e.Logf("Instances state should be terminated, actual state is %s", reservation.Instances[0].State.Name)
return false
}
}
return true
}
o.Eventually(waitUntilSpotVMTerminated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
exutil.By("Start cluster again, check ClusterDeployment back to running again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
expectKeyValue2 := map[string]string{
"status": "False",
"reason": "ResumingOrRunning",
"message": "Cluster is resuming or running, see Ready condition for details",
}
waitForHibernating2 := checkCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "Hibernating", expectKeyValue2, "wait for cluster being resumed")
o.Eventually(waitForHibernating2).WithTimeout(10 * time.Minute).WithPolling(15 * time.Second).Should(o.BeTrue())
e2e.Logf("Making sure the cluster is in the \"Running\" powerstate ...")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Login to target cluster, check the new spot instances are created")
var newSpotMachineName, newSpotMachineName2 string
checkSpotMachineName2 := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("--kubeconfig="+kubeconfig, "machine", "-n", "openshift-machine-api", "-o=jsonpath={.items[*].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("machine list: %s", stdout)
newSpotMachineName = ""
newSpotMachineName2 = ""
for _, MachineName := range strings.Split(stdout, " ") {
if strings.Contains(MachineName, "spot-") {
newSpotMachineName = MachineName
}
if strings.Contains(MachineName, "spot2-") {
newSpotMachineName2 = MachineName
}
}
e2e.Logf("newSpotMachineName: %s, newSpotMachineName2: %s", newSpotMachineName, newSpotMachineName2)
return strings.Contains(newSpotMachineName, "spot-") && strings.Contains(newSpotMachineName2, "spot2-") && oldSpotMachineName != newSpotMachineName && oldSpotMachineName2 != newSpotMachineName2
}
o.Eventually(checkSpotMachineName2).WithTimeout(DefaultTimeout * time.Second).WithPolling(5 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 058485e5-ffe0-4deb-843a-5b8ba9b47ec4 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-32135-kubeconfig and password secrets need to be owned by ClusterDeployment after installed [Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-32135-kubeconfig and password secrets need to be owned by ClusterDeployment after installed [Serial]", func() {
testCaseID := "32135"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check ownerReference for secrets kubeconfig and password, before installed, it is only owned by ClusterProvision.")
ClusterprovisionName := getClusterprovisionName(oc, cdName, oc.Namespace())
kubeconfigName := ClusterprovisionName + "-admin-kubeconfig"
passwordName := ClusterprovisionName + "-admin-password"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", nok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", nok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check ownership again, it will be owned by both ClusterProvision and ClusterDeployment.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Delete ClusterProvision.")
err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterProvision", ClusterprovisionName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check kubeconfig and password secrets are still exist and owned by clusterdeployment.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", nok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", kubeconfigName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterProvision", nok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterDeployment", ok, DefaultTimeout, []string{"secret", passwordName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.ownerReferences}"}).check(oc)
exutil.By("Delete clusterdeployment, kubeconfig and password secrets will be deleted.")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, kubeconfigName, nok, DefaultTimeout, []string{"secret", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, passwordName, nok, DefaultTimeout, []string{"secret", "-n", oc.Namespace()}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 351a7df5-2104-4031-be2e-3884a644102a | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-43029-Hive should abandon deprovision when preserveOnDelete is true when clusters with managed DNS [Serial] | ['"encoding/json"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-High-43029-Hive should abandon deprovision when preserveOnDelete is true when clusters with managed DNS [Serial]", func() {
testCaseID := "43029"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create Route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, FakeClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Edit secret aws-creds and change the data to an invalid value")
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "--type", `merge`, `--patch={"data": {"aws_access_key_id": "MTIzNDU2"}}`, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Delete the cd, and then hive will hit DeprovisionLaunchError=AuthenticationFailed, and stuck in deprovision process")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
waitForDeprovisionLaunchError := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "DeprovisionLaunchError")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition DeprovisionLaunchError, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "AuthenticationFailed" {
e2e.Logf("For condition DeprovisionLaunchError, expected reason is AuthenticationFailed, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "Credential check failed") != 0 {
e2e.Logf("For condition DeprovisionLaunchError, expected message is \nCredential check failed, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition DeprovisionLaunchError, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForDeprovisionLaunchError).WithTimeout(ClusterUninstallTimeout * time.Second).WithPolling(30 * time.Second).Should(o.BeTrue())
exutil.By("Set cd.spec.preserveOnDelete = true on cd")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "--type", "json", "-p", "[{\"op\": \"remove\", \"path\": \"/spec/preserveOnDelete\"}]").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "--type", `merge`, `--patch={"spec": {"preserveOnDelete": true}}`, "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check in this situation, hive would be able to remove dnszone and CD CR directly")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 8f9a211f-6c38-4aaa-932e-e91c696103f3 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Critical-28631-[HiveSpec] Hive deprovision controller can be disabled through a hiveconfig option [Serial] | ['"encoding/json"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Critical-28631-[HiveSpec] Hive deprovision controller can be disabled through a hiveconfig option [Serial]", func() {
testCaseID := "28631"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
e2e.Logf("Wait until infra id generated")
newCheck("expect", "get", asAdmin, false, contain, cdName+"-", ok, 600, []string{"cd", "-n", oc.Namespace()}).check(oc)
oldhivecontrollersPod := getHivecontrollersPod(oc, HiveNamespace)
e2e.Logf("old hivecontrollers Pod is " + oldhivecontrollersPod)
e2e.Logf("Add \"deprovisionsDisabled: true\" in hiveconfig.spec")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", "json", "-p", "[{\"op\": \"remove\", \"path\": \"/spec/deprovisionsDisabled\"}]").Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"deprovisionsDisabled": true}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check \"deprovisionsDisabled\" is set to true in hiveconfig.spec")
newCheck("expect", "get", asAdmin, false, compare, "true", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.deprovisionsDisabled}"}).check(oc)
e2e.Logf("Check if hivecontrollers Pod is recreated")
var hivecontrollersPod string
checkNewcontrollersPod := func() bool {
hivecontrollersPod = getHivecontrollersPod(oc, HiveNamespace)
return strings.Compare(oldhivecontrollersPod, hivecontrollersPod) != 0
}
o.Eventually(checkNewcontrollersPod).WithTimeout(120 * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("new hivecontrollers Pod is " + hivecontrollersPod)
e2e.Logf("Try to delete cd")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
e2e.Logf(`Check logs of hive-controllers has a warning :"deprovisions are currently disabled in HiveConfig, skipping"`)
checkDeprovisionLog := func() bool {
deprovisionLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hivecontrollersPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(deprovisionLogs, "deprovisions are currently disabled in HiveConfig, skipping") {
e2e.Logf(`Find target message :"deprovisions are currently disabled in HiveConfig, skipping"`)
return true
}
e2e.Logf(`Still waiting for message :"deprovisions are currently disabled in HiveConfig, skipping"`)
return false
}
o.Eventually(checkDeprovisionLog).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
e2e.Logf("Add \"deprovisionsDisabled: false\" in hiveconfig.spec")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"deprovisionsDisabled": false}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check \"deprovisionsDisabled\" is set to false in hiveconfig.spec")
newCheck("expect", "get", asAdmin, false, compare, "false", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.deprovisionsDisabled}"}).check(oc)
e2e.Logf("Check if cd is in deprovision.")
newCheck("expect", "get", asAdmin, false, contain, cdName+"-uninstall-", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace()}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 8b746516-a74e-46e3-805a-dbf9f980494b | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-41286-ClusterPool supports provisioning fake cluster [Serial] | ['"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:sguo-Medium-41286-ClusterPool supports provisioning fake cluster [Serial]", func() {
testCaseID := "41286"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
exutil.By("Check hive will propagate the annotation to all created ClusterDeployment")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
newCheck("expect", "get", asAdmin, withoutNamespace, contain, `"hive.openshift.io/fake-cluster":"true"`, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.metadata.annotations}"}).check(oc)
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check claiming a fake cluster works well")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
waitForClusterClaimRunning := func() bool {
condition := getCondition(oc, "ClusterClaim", claimName, oc.Namespace(), "ClusterRunning")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition ClusterRunning, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Running" {
e2e.Logf("For condition ClusterRunning, expected reason is Running, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || message != "Cluster is running" {
e2e.Logf("For condition ClusterRunning, expected message is \nCluster is running, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition ClusterRunning, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClusterClaimRunning).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check clusterMetadata field of fake cluster, all fields have values")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.adminKubeconfigSecretRef.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.adminPasswordSecretRef.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.spec.clusterMetadata.infraID}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | ec18e89e-ca0b-42ad-8cfc-c80a5c6a4c26 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-42661-Low-39179-Simulate hibernation for fake clusters [Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Medium-42661-Low-39179-Simulate hibernation for fake clusters [Serial]", func() {
testCaseID := "42661"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create fake ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "1m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Check all clusters in cluster pool are in Hibernating status")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check claiming a fake cluster works well")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("Check cluster is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Hibernating it again, check it can be hibernated again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", cdNameSpace, "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.powerState}"}).check(oc)
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create fake ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
installConfigSecret.create(oc)
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("Check fake cluster is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Hibernating the fake cluster ,check it can be hibernated")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Hibernating"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Hibernating", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Restart it again, check it back to running again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", `--patch={"spec":{"powerState": "Running"}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, HibernateAfterTimer, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("OCP-39179: Autogenerate Cluster ID in hive for fake cluster instead of fixed value fake-cluster-id")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "fake-cluster-id", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "fake-cluster-", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterMetadata.clusterID}"}).check(oc)
ClusterprovisionName := getClusterprovisionName(oc, cdName, oc.Namespace())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "fake-cluster-id", nok, DefaultTimeout, []string{"ClusterProvision", ClusterprovisionName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterID}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "fake-cluster-", ok, DefaultTimeout, []string{"ClusterProvision", ClusterprovisionName, "-n", oc.Namespace(), "-o=jsonpath={.spec.clusterID}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | ce69365f-870d-4b7f-8bac-4e962d50bd98 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Low-25443-Low-29855-Clusterdeployment contains Status.Condition of SyncSet status in case of syncset is invalid [Serial] | ['"fmt"', '"os"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:sguo-Low-25443-Low-29855-Clusterdeployment contains Status.Condition of SyncSet status in case of syncset is invalid [Serial]", func() {
testCaseID := "25443"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create SyncSet for resource apply......")
syncSetName := testCaseID + "-syncset1"
configMapName := testCaseID + "-configmap1"
configMapNamespace := testCaseID + "-configmap1-ns"
resourceMode := "Sync"
syncTemp := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource := syncSetResource{
name: syncSetName,
namespace: oc.Namespace(),
namespace2: configMapNamespace,
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: resourceMode,
template: syncTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName})
syncResource.create(oc)
e2e.Logf("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Check if SyncSetPatch is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName, ok, DefaultTimeout, []string{"SyncSet", syncSetName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if Syncset is not failed before applying the patch.")
waitForSyncsetSuccess := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "SyncSetFailed")
if status, ok := condition["status"]; !ok || status != "False" {
e2e.Logf("For condition SyncSetFailed, expected status is False, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "SyncSetApplySuccess" {
e2e.Logf("For condition SyncSetFailed, expected reason is SyncSetApplySuccess, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "SyncSet apply is successful") != 0 {
e2e.Logf("For condition SyncSetFailed, expected message is \nSyncSet apply is successful, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition SyncSetFailed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForSyncsetSuccess).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
syncSetPatchName := testCaseID + "-syncset-patch"
syncPatchTemp := filepath.Join(testDataDir, "syncset-patch.yaml")
patchContent := ` { "data": { "foo": "new-bar" }`
patchType := "merge"
syncPatch := syncSetPatch{
name: syncSetPatchName,
namespace: oc.Namespace(),
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
pcontent: patchContent,
patchType: patchType,
template: syncPatchTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
syncPatch.create(oc)
e2e.Logf("Check if Syncset is failed.")
waitForSyncsetFail := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "SyncSetFailed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition SyncSetFailed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "SyncSetApplyFailure" {
e2e.Logf("For condition SyncSetFailed, expected reason is SyncSetApplyFailure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, "One of the SyncSet applies has failed") != 0 {
e2e.Logf("For condition SyncSetFailed, expected message is \nOne of the SyncSet applies has failed, \nactual reason is %v\n, retrying ...", message)
return false
}
e2e.Logf("For condition SyncSetFailed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForSyncsetFail).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-29855:Hive treates bad syncsets as controller warnings instead of controller errors")
waitForClustersyncFail1 := func() bool {
condition := getCondition(oc, "clustersync", cdName, oc.Namespace(), "Failed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition Failed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Failure" {
e2e.Logf("For condition Failed, expected reason is Failure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, fmt.Sprintf("SyncSet %s is failing", syncSetPatchName)) != 0 {
e2e.Logf("For condition Failed, expected message is \nSyncSet %v is failing, \nactual reason is %v\n, retrying ...", syncSetPatchName, message)
return false
}
e2e.Logf("For condition Failed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClustersyncFail1).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
hiveclustersyncPod := "hive-clustersync-0"
e2e.Logf(`Check logs of hive-clustersync-0 has a warning log instead of error log`)
checkclustersyncLog1 := func() bool {
clustersyncLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveclustersyncPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(clustersyncLogs, "level=warning msg=\"running the patch command failed\"") {
e2e.Logf(`Find target message :level=warning msg="running the patch command failed"`)
return true
}
e2e.Logf(`Still waiting for message :level=warning msg="running the patch command failed"`)
return false
}
o.Eventually(checkclustersyncLog1).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
exutil.By("Extracting kubeconfig ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
syncSetSecretName := testCaseID + "-syncset-secret"
syncSecretTemp := filepath.Join(testDataDir, "syncset-secret.yaml")
sourceName := testCaseID + "-secret"
syncSecret := syncSetSecret{
name: syncSetSecretName,
namespace: oc.Namespace(),
cdrefname: cdName,
sname: "secret-not-exist",
snamespace: oc.Namespace(),
tname: sourceName,
tnamespace: "default",
template: syncSecretTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetSecretName})
syncSecret.create(oc)
e2e.Logf("Check if Syncset-secret failed to apply.")
waitForClustersyncFail2 := func() bool {
condition := getCondition(oc, "clustersync", cdName, oc.Namespace(), "Failed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition Failed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "Failure" {
e2e.Logf("For condition Failed, expected reason is Failure, actual reason is %v, retrying ...", reason)
return false
}
if message, ok := condition["message"]; !ok || strings.Compare(message, fmt.Sprintf("SyncSet %s is failing", syncSetSecretName)) != 0 {
e2e.Logf("For condition Failed, expected message is \nSyncSet %v is failing, \nactual reason is %v\n, retrying ...", syncSetSecretName, message)
return false
}
e2e.Logf("For condition Failed, fields status, reason & message all expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForClustersyncFail2).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("Check target cluster doesn't have this secret.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "secret"}).check(oc)
e2e.Logf(`Check logs of hive-clustersync-0 doesn't have error log`)
checkclustersyncLog2 := func() bool {
clustersyncLogs, _, err := oc.AsAdmin().WithoutNamespace().Run("logs").Args(hiveclustersyncPod, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
if strings.Contains(clustersyncLogs, fmt.Sprintf("level=info msg=\"cannot read secret\" SyncSet=%s", syncSetSecretName)) {
e2e.Logf(`Find target message :level=info msg="cannot read secret"`)
return true
}
e2e.Logf(`Still waiting for message :level=info msg="cannot read secret"`)
return false
}
o.Eventually(checkclustersyncLog2).WithTimeout(600 * time.Second).WithPolling(60 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | afb5136c-c85a-4b3e-8f35-7228177e5319 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Critical-23986-Medium-64550-[HiveSpec] Kubeconfig secrets can work with additional CAs[Serial] | ['"crypto/ecdsa"', '"crypto/elliptic"', '"crypto/rand"', '"crypto/tls"', '"encoding/json"', '"fmt"', '"os"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/go-acme/lego/v4/certificate"', '"github.com/go-acme/lego/v4/lego"', '"github.com/go-acme/lego/v4/registration"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Critical-23986-Medium-64550-[HiveSpec] Kubeconfig secrets can work with additional CAs[Serial]", func() {
testCaseID := "23986"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
apiEndpoint := "api." + cdName + "." + AWSBaseDomain
appsEndpoint := "apps." + cdName + "." + AWSBaseDomain
appsEndpointGlobbing := "*." + appsEndpoint
appsEndpointConsole := "console-openshift-console." + appsEndpoint
/*
To generate a Let's Encrypt certificate, we have the following options:
1) Use the cert-manager operator:
Pro: Openshift native
Con: we are no longer testing Hive itself as we rely on another operator as well
2) Use certbot (or hiveutil which relies on it):
Pro: straightforwardness
Con: we have to install certbot
3) Use a Golang library which automates this process:
Pro: straightforwardness (somewhat)
Con: cannot think of any
Here we are using option 3).
*/
exutil.By("Getting a Let's Encrypt certificate for " + apiEndpoint + " & " + appsEndpointGlobbing)
// Get Lego user and config
privateKey, err := ecdsa.GenerateKey(elliptic.P256(), rand.Reader)
o.Expect(err).NotTo(o.HaveOccurred())
user := legoUser{key: privateKey}
config := lego.NewConfig(&user)
// Get Lego client
client, err := lego.NewClient(config)
o.Expect(err).NotTo(o.HaveOccurred())
// Registration for new user
_, err = client.Registration.Register(registration.RegisterOptions{TermsOfServiceAgreed: true})
o.Expect(err).NotTo(o.HaveOccurred())
// Set Lego DNS provider which is used to solve the ACME DNS challenge
// (and cleanup the related DNS records after that)
maxRetries := 5
TTL := 10
propagationTimeout, pollingInterval := 15*time.Minute, 4*time.Second
awsAccessKeyId, awsSecretAccessKey := getAWSCredentials(oc)
dnsProvider, err := newLegoDNSProvider(maxRetries, TTL, propagationTimeout, pollingInterval, awsAccessKeyId, awsSecretAccessKey, AWSRegion)
o.Expect(err).NotTo(o.HaveOccurred())
err = client.Challenge.SetDNS01Provider(dnsProvider)
o.Expect(err).NotTo(o.HaveOccurred())
// Request for certificates
// Note:
// Lego checks DNS record propagation from recursive DNS servers specified in /etc/resolv.conf (if possible).
// So before running this test case locally, turn off the VPNs as they often update /etc/resolv.conf.
request := certificate.ObtainRequest{
Domains: []string{apiEndpoint, appsEndpointGlobbing},
// We want the certificates to be split
Bundle: false,
}
certificates, err := client.Certificate.Obtain(request)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Writing certificates & private key to files...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
fullChainFilePath := tmpDir + "/fullchain.pem"
err = os.WriteFile(fullChainFilePath, append(certificates.Certificate, certificates.IssuerCertificate...), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
chainFilePath := tmpDir + "/chain.pem"
err = os.WriteFile(chainFilePath, certificates.IssuerCertificate, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
privateKeyFilePath := tmpDir + "/privkey.pem"
err = os.WriteFile(privateKeyFilePath, certificates.PrivateKey, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating serving-cert Secret which will be referenced in CD's manifest...")
servingCertificateSecretName := "serving-cert"
defer oc.AsAdmin().Run("delete").Args("secret", servingCertificateSecretName).Execute()
err = oc.AsAdmin().Run("create").Args("secret", "tls", servingCertificateSecretName, "--cert="+fullChainFilePath, "--key="+privateKeyFilePath).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ca-cert Secret which will be referenced in HiveConfig/hive...")
caCertificateSecretName := "ca-cert"
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", caCertificateSecretName, "-n=hive").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", caCertificateSecretName, "--from-file=ca.crt="+chainFilePath, "-n=hive").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Referencing ca-cert Secret in HiveConfig/hive...")
patch := `
spec:
additionalCertificateAuthoritiesSecretRef:
- name: ` + caCertificateSecretName
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/additionalCertificateAuthoritiesSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterDeployment...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cd := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, cd.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cd.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cd)
exutil.By("Patching CD s.t. it references the serving certificate Secret...")
patch = fmt.Sprintf(`
spec:
certificateBundles:
- name: serving-cert
certificateSecretRef:
name: %s
controlPlaneConfig:
servingCertificates:
default: serving-cert
ingress:
- name: default
domain: %s
servingCertificate: serving-cert`, servingCertificateSecretName, appsEndpoint)
err = oc.AsAdmin().Run("patch").Args("clusterdeployment", cdName, "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Waiting for the CD to be installed...")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Making sure the target cluster is using the right certificate...")
endpointCertIsGood := func(endpoint string) bool {
e2e.Logf("Checking certificates for endpoint %v ...", endpoint)
conn, err := tls.Dial("tcp", endpoint, &tls.Config{InsecureSkipVerify: true})
if err != nil {
e2e.Logf("Error dialing endpoint %v: %v, keep polling ...", endpoint, err.Error())
return false
}
// Must call conn.Close() here to make sure the connection is successfully established,
// so the conn object is populated and can be closed without incurring a nil pointer dereference error.
defer conn.Close()
// Look for the target certificate (the one with apiEndpoint/appsEndpoint as subject)
// in all certificates of the endpoint
for _, cert := range conn.ConnectionState().PeerCertificates {
if strings.Contains(cert.Subject.String(), apiEndpoint) || strings.Contains(cert.Subject.String(), appsEndpoint) {
// For simplicity, here we only check the issuer is correct on the target certificate
return strings.Contains(cert.Issuer.String(), `Let's Encrypt`)
}
}
e2e.Logf("Target certificate not found on endpoint %v, keep polling ...", endpoint)
return false
}
// It seems that DNS propagation can be really slow for "*.apps.CLUSTER.qe.devcluster.openshift.com" (literally)
// So here we check the console endpoint "console.apps.CLUSTER.qe.devcluster.openshift.com" instead
checkCertificates := func() bool {
return endpointCertIsGood(apiEndpoint+":6443") && endpointCertIsGood(appsEndpointConsole+":443")
}
// We need to poll s.t. remote-ingress or control-plane-certificate-related SyncSets are applied
// and APIServer/Ingress-Operator finish reconcile on the target cluster.
o.Eventually(checkCertificates).WithTimeout(20 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
// The kubeconfig obtained (for ex. Secret/fxie-hive-1-0-wlqg2-admin-kubeconfig.data["kubeconfig"]) has the
// CA certs integrated, so we should be able to communicate to the target cluster without the following error:
// "x509: certificate signed by unknown authority".
exutil.By("Communicating to the target cluster using the kubeconfig with Let's Encrypt's CA...")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfigPath := tmpDir + "/kubeconfig"
err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "--kubeconfig", kubeconfigPath).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("OCP-64550: Hive should be able to delete Secret/hive-additional-ca")
// Make sure the hive-additional-CA Secret still exists at this moment
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Secret", hiveAdditionalCASecret, "-n", HiveNamespace).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(stdout).To(o.ContainSubstring(hiveAdditionalCASecret))
// Patch HiveConfig
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/additionalCertificateAuthoritiesSecretRef"}]`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
// Make sure the hive-additional-CA Secret is eventually deleted
hiveOperatorReconcileTimeout := 300
newCheck("expect", "get", asAdmin, withoutNamespace, contain, hiveAdditionalCASecret, nok, hiveOperatorReconcileTimeout, []string{"Secret", "-n", HiveNamespace}).check(oc)
// Make sure Hive Operator stays healthy for a while
hiveIsStillHealthy := func() bool {
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("hiveconfig/hive", `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
return stdout == "True"
}
o.Consistently(hiveIsStillHealthy).WithTimeout(DefaultTimeout * time.Second).WithPolling(10 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 2a3092e1-9fd9-4113-9017-5fd55ab2647c | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25145-[HiveSpec] Dynamically detect change to global pull secret content [Serial] | ['"encoding/json"', '"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25145-[HiveSpec] Dynamically detect change to global pull secret content [Serial]", func() {
testCaseID := "25145"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Preparing an incomplete pull-secret ...")
var pullSecretMapIncomplete map[string]map[string]map[string]string
stdout, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/pull-secret", "-n", "openshift-config", "--to", "-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &pullSecretMapIncomplete)
o.Expect(err).NotTo(o.HaveOccurred())
delete(pullSecretMapIncomplete["auths"], "registry.ci.openshift.org")
exutil.By("Creating an incomplete pull-secret in Hive's namespace and the temporary project's namespace respectively ...")
pullSecretBsIncomplete, _ := json.Marshal(pullSecretMapIncomplete)
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", PullSecret, "-n", HiveNamespace).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", PullSecret, "--from-literal=.dockerconfigjson="+string(pullSecretBsIncomplete), "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.Run("delete").Args("secret", PullSecret).Execute()
err = oc.Run("create").Args("secret", "generic", PullSecret, "--from-literal=.dockerconfigjson="+string(pullSecretBsIncomplete)).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Patching HiveConfig so that it refers to an incomplete global pull-secret ...")
patch := `
spec:
globalPullSecretRef:
name: ` + PullSecret
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/globalPullSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterImageSet ...")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret ...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Copying AWS credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment with an incomplete pull-secret ...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("Waiting for the cluster installation to fail ...")
waitForAPIWaitFailure := func() bool {
condition := getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "ProvisionFailed")
if status, ok := condition["status"]; !ok || status != "True" {
e2e.Logf("For condition ProvisionFailed, expected status is True, actual status is %v, retrying ...", status)
return false
}
if reason, ok := condition["reason"]; !ok || reason != "KubeAPIWaitFailed" {
e2e.Logf("For condition ProvisionFailed, expected reason is KubeAPIWaitFailed, actual reason is %v, retrying ...", reason)
return false
}
e2e.Logf("For condition ProvisionFailed, fields status and reason expected, proceeding to the next step ...")
return true
}
o.Eventually(waitForAPIWaitFailure).WithTimeout(ClusterInstallTimeout * time.Second).WithPolling(3 * time.Minute).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 971ac639-1b23-4a37-9d80-183f73882e74 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25210-Collect ClusterOperator Status for Hive Managed Clusters [Serial] | ['"encoding/json"', '"os"', '"path/filepath"', '"reflect"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-25210-Collect ClusterOperator Status for Hive Managed Clusters [Serial]", func() {
testCaseID := "25210"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating install-config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Making sure the cluster is installed and in the \"Running\" powerstate ...")
newCheck("expect", "get", asAdmin, false, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, false, compare, "Running", ok, WaitingForClusterOperatorsTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Extracting kubeconfig ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
exutil.By("Comparing conditions obtained from ClusterOperator and ClusterState ...")
var clusterStateConditions, clusterOperatorConditions map[string][]map[string]string
clusterStateJSONPath := `{"{"}{range .status.clusterOperators[:-1]}"{.name}":{.conditions},{end}{range .status.clusterOperators[-1]}"{.name}":{.conditions}{end}{"}"}`
clusterOperatorJSONPath := `{"{"}{range .items[:-1]}"{.metadata.name}":{.status.conditions},{end}{range .items[-1]}"{.metadata.name}":{.status.conditions}{end}{"}"}`
/*
stdout[any-index] =
{
"operator-name": [
{
"lastTransitionTime": ...
...
}
]
}
*/
checkConditionEquality := func() bool {
stdout, _, err := oc.AsAdmin().Run("get").Args("ClusterState", cdName, "-o", "jsonpath="+clusterStateJSONPath).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &clusterStateConditions)
o.Expect(err).NotTo(o.HaveOccurred())
stdout, _, err = oc.AsAdmin().Run("get").Args("ClusterOperator", "-o", "jsonpath="+clusterOperatorJSONPath, "--kubeconfig="+kubeconfig).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
err = json.Unmarshal([]byte(stdout), &clusterOperatorConditions)
o.Expect(err).NotTo(o.HaveOccurred())
return reflect.DeepEqual(clusterOperatorConditions, clusterStateConditions)
}
o.Eventually(checkConditionEquality).WithTimeout(20 * time.Minute).WithPolling(time.Minute).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | e86fd2f3-5281-4dc8-8847-23073f1d5f18 | Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-33832-Low-42251-Medium-43033-[HiveSpec] Hive supports ClusterPool [Serial] | ['"path/filepath"', '"reflect"', '"sort"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/credentials"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-33832-Low-42251-Medium-43033-[HiveSpec] Hive supports ClusterPool [Serial]", func() {
testCaseID := "33832"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 1,
maxSize: 1,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "360m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterPool", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-42251 - Initialize hive CR conditions")
exutil.By("OCP-42251 Step 1: Check all conditions type of ClusterPool")
allClusterPoolConditionTypes := []string{"MissingDependencies", "CapacityAvailable", "AllClustersCurrent", "InventoryValid", "DeletionPossible"}
sort.Strings(allClusterPoolConditionTypes)
checkClusterPoolConditionType := func() bool {
checkedClusterPoolConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterPoolConditionTypes := strings.Split(checkedClusterPoolConditionTypesOutput, " ")
sort.Strings(checkedClusterPoolConditionTypes)
e2e.Logf("Compare allClusterPoolConditionTypes: %v and checkedClusterPoolConditionTypes: %v", allClusterPoolConditionTypes, checkedClusterPoolConditionTypes)
return reflect.DeepEqual(allClusterPoolConditionTypes, checkedClusterPoolConditionTypes)
}
o.Eventually(checkClusterPoolConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
e2e.Logf("Check if ClusterDeployment is created")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, poolName, ok, DefaultTimeout, []string{"ClusterDeployment", "-A"}).check(oc)
exutil.By("OCP-42251 Step 2: Check all conditions type of ClusterDeployment")
cdName, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", "-A", "-o=jsonpath={.items[0].metadata.name}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
cdNameSpace := cdName
allClusterDeploymentConditionTypes := []string{"InstallerImageResolutionFailed", "ControlPlaneCertificateNotFound", "IngressCertificateNotFound", "Unreachable", "ActiveAPIURLOverride",
"DNSNotReady", "InstallImagesNotResolved", "ProvisionFailed", "SyncSetFailed", "RelocationFailed", "Hibernating", "Ready", "InstallLaunchError", "DeprovisionLaunchError", "ProvisionStopped",
"Provisioned", "RequirementsMet", "AuthenticationFailure", "AWSPrivateLinkReady", "AWSPrivateLinkFailed", "ClusterInstallFailed", "ClusterInstallCompleted", "ClusterInstallStopped", "ClusterInstallRequirementsMet"}
sort.Strings(allClusterDeploymentConditionTypes)
checkClusterDeploymentConditionType := func() bool {
checkedClusterDeploymentConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", cdNameSpace, "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterDeploymentConditionTypes := strings.Split(checkedClusterDeploymentConditionTypesOutput, " ")
sort.Strings(checkedClusterDeploymentConditionTypes)
e2e.Logf("Compare allClusterDeploymentConditionTypes: %v and checkedClusterDeploymentConditionTypes: %v", allClusterDeploymentConditionTypes, checkedClusterDeploymentConditionTypes)
return reflect.DeepEqual(allClusterDeploymentConditionTypes, checkedClusterDeploymentConditionTypes)
}
o.Eventually(checkClusterDeploymentConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("OCP-42251 Step 3: Check all conditions type of MachinePool")
machinepoolName := cdName + "-worker"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, machinepoolName, ok, DefaultTimeout, []string{"Machinepool", "-n", cdNameSpace}).check(oc)
allMachinepoolConditionTypes := []string{"NotEnoughReplicas", "NoMachinePoolNameLeasesAvailable", "InvalidSubnets", "UnsupportedConfiguration"}
sort.Strings(allMachinepoolConditionTypes)
checkMachinePoolConditionType := func() bool {
checkedMachinepoolConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("Machinepool", machinepoolName, "-n", cdNameSpace, "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedMachinepoolConditionTypes := strings.Split(checkedMachinepoolConditionTypesOutput, " ")
sort.Strings(checkedMachinepoolConditionTypes)
e2e.Logf("Compare allMachinepoolConditionTypes: %v and checkedMachinepoolConditionTypes: %v", allMachinepoolConditionTypes, checkedMachinepoolConditionTypes)
return reflect.DeepEqual(allMachinepoolConditionTypes, checkedMachinepoolConditionTypes)
}
o.Eventually(checkMachinePoolConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check if ClusterPool become ready")
//runningCount is 0 so pool status should be standby: 1, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, ClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
exutil.By("Create ClusterClaim...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
exutil.By("Check if ClusterClaim created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, claimName, ok, DefaultTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-42251 Step 4: Check all conditions type of ClusterClaim")
allClusterClaimConditionTypes := []string{"Pending", "ClusterRunning"}
sort.Strings(allClusterClaimConditionTypes)
checkClusterClaimConditionType := func() bool {
checkedClusterClaimConditionTypesOutput, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterClaim", claimName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[*].type}").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
checkedClusterClaimConditionTypes := strings.Split(checkedClusterClaimConditionTypesOutput, " ")
sort.Strings(checkedClusterClaimConditionTypes)
e2e.Logf("Compare allClusterClaimConditionTypes: %v and checkedClusterClaimConditionTypes: %v", allClusterClaimConditionTypes, checkedClusterClaimConditionTypes)
return reflect.DeepEqual(allClusterClaimConditionTypes, checkedClusterClaimConditionTypes)
}
o.Eventually(checkClusterClaimConditionType).WithTimeout(DefaultTimeout * time.Second).WithPolling(3 * time.Second).Should(o.BeTrue())
exutil.By("Check if ClusterClaim become running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterClaim", "-n", oc.Namespace()}).check(oc)
exutil.By("OCP-43033: oc get clusterclaim should report ClusterDeleted")
exutil.By("Delete the ClusterDeployment")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", cdNameSpace).Background()
defer cmd.Process.Kill()
exutil.By("Check ClusterRunning conditions of clusterclaim")
expectKeyValue := map[string]string{
"status": "False",
"reason": "ClusterDeleted",
"message": "Assigned cluster has been deleted",
}
waitForClusterRunningFalse := checkCondition(oc, "ClusterClaim", claimName, oc.Namespace(), "ClusterRunning", expectKeyValue, "wait for ClusterRunning false")
o.Eventually(waitForClusterRunningFalse).WithTimeout(ClusterUninstallTimeout * time.Second).WithPolling(15 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | db606864-0152-4197-8dc1-bae915700e6f | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-23167-Critical-73287-The tags created on users in AWS match what the installer did on your instances [Serial] | ['"context"', '"os"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"github.com/aws/aws-sdk-go-v2/service/iam"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-23167-Critical-73287-The tags created on users in AWS match what the installer did on your instances [Serial]", func() {
testCaseID := "23167"
cdName := "cd-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating ClusterDeployment ...")
installConfig := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cd := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cd.name+"-imageset", oc.Namespace(), installConfig.name1, cd.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfig, cd)
exutil.By("Create the infra MachinePool ...")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
inframp.create(oc)
// Wait for the cluster to be installed and extract its infra id
newCheck("expect", "get", asAdmin, false, compare, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
infraID, _, err := oc.AsAdmin().Run("get").Args("cd", cdName, "-o", "jsonpath='{.spec.clusterMetadata.infraID}'").Outputs()
infraID = strings.Trim(infraID, "'")
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Cluster infraID: " + infraID)
// Extract AWS credentials
AWSAccessKeyID, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/aws-creds", "-n=kube-system", "--keys=aws_access_key_id", "--to=-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
AWSSecretAccessKey, _, err := oc.AsAdmin().WithoutNamespace().Run("extract").Args("secret/aws-creds", "-n=kube-system", "--keys=aws_secret_access_key", "--to=-").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
// AWS clients
cfg, err := config.LoadDefaultConfig(
context.Background(),
config.WithCredentialsProvider(credentials.NewStaticCredentialsProvider(AWSAccessKeyID, AWSSecretAccessKey, "")),
config.WithRegion(AWSRegion),
)
o.Expect(err).NotTo(o.HaveOccurred())
ec2Client := ec2.NewFromConfig(cfg)
iamClient := iam.NewFromConfig(cfg)
// Make sure resources are created with the target tag
targetTag := "kubernetes.io/cluster/" + infraID
exutil.By("Checking that resources are created with the target tag " + targetTag)
describeTagsOutput, err := ec2Client.DescribeTags(context.Background(), &ec2.DescribeTagsInput{
Filters: []types.Filter{
{
Name: aws.String("key"),
Values: []string{targetTag},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeTagsOutput.Tags)).NotTo(o.BeZero())
// Make sure the IAM users are tagged
exutil.By("Looking for IAM users prefixed with infraID ...")
pagination := aws.Int32(50)
userFound, username := false, ""
listUsersOutput := &iam.ListUsersOutput{}
err = wait.Poll(6*time.Second, 10*time.Minute, func() (bool, error) {
listUsersOutput, err = iamClient.ListUsers(context.Background(), &iam.ListUsersInput{
Marker: listUsersOutput.Marker,
MaxItems: pagination,
})
o.Expect(err).NotTo(o.HaveOccurred())
for _, user := range listUsersOutput.Users {
if strings.HasPrefix(*user.UserName, infraID) {
userFound, username = true, *user.UserName
break
}
}
if userFound {
return true, nil
}
return false, nil
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(userFound).To(o.BeTrue())
exutil.By("Looking for tags on user " + username)
listUserTagsOutput, err := iamClient.ListUserTags(context.Background(), &iam.ListUserTagsInput{
UserName: aws.String(username),
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(*listUserTagsOutput.Tags[0].Key).To(o.Equal(targetTag))
o.Expect(*listUserTagsOutput.Tags[0].Value).To(o.Equal("owned"))
exutil.By("Check OCP-73287, the machine in infra MP shall have userTags defined in both CD and MP")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
instanceId := getResource(oc, asAdmin, withoutNamespace, "machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[0].status.providerStatus.instanceId}", "--kubeconfig="+kubeconfig)
e2e.Logf("instanceId is %v", instanceId)
describeTagsOutput, err = ec2Client.DescribeTags(context.Background(), &ec2.DescribeTagsInput{
Filters: []types.Filter{
{
Name: aws.String("resource-id"),
Values: []string{instanceId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
tagsLen := len(describeTagsOutput.Tags)
e2e.Logf("Tags length = %v", tagsLen)
numMatchedTags := 0
for i := range describeTagsOutput.Tags {
e2e.Logf("userTag %v, Key = %v, Value = %v", i, *describeTagsOutput.Tags[i].Key, *describeTagsOutput.Tags[i].Value)
// Tag defined in CD
if *describeTagsOutput.Tags[i].Key == "hive-qe-tag1" && *describeTagsOutput.Tags[i].Value == "hive-qe-value1" {
numMatchedTags++
}
// Tag defined in MP
if *describeTagsOutput.Tags[i].Key == "hive-qe-tag3" && *describeTagsOutput.Tags[i].Value == "hive-qe-value3" {
numMatchedTags++
}
// Tag defined in both CD and MP, MP value shall take precedence
if *describeTagsOutput.Tags[i].Key == "hive-qe-customizedtag" && *describeTagsOutput.Tags[i].Value == AWSDefaultMPTag {
numMatchedTags++
}
}
//Totally 3 tags matched
o.Expect(numMatchedTags).To(o.Equal(3))
}) | |||||
test case | openshift/openshift-tests-private | 1ee952bd-bbff-43e9-a263-03994bc4d5b9 | Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-25310-High-33374-High-39747-Medium-23165-High-22760- [HiveSDRosa] Hive ClusterDeployment Check installed and version [Serial] | ['"os"', '"path/filepath"', '"strconv"', '"strings"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-25310-High-33374-High-39747-Medium-23165-High-22760- [HiveSDRosa] Hive ClusterDeployment Check installed and version [Serial]", func() {
testCaseID := "25310"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Selecting a custom OCP version to install ...")
ocpVersion := extractRelFromImg(testOCPImage)
xyzVersion := strings.Split(ocpVersion, ".")
majorVersion := xyzVersion[0]
minorVersion := xyzVersion[1]
patchVersion := xyzVersion[2]
minorVersionInt, err := strconv.Atoi(minorVersion)
o.Expect(err).NotTo(o.HaveOccurred())
minorVersion = strconv.Itoa(minorVersionInt - 1)
customOCPImage, err := exutil.GetLatestNightlyImage(majorVersion + "." + minorVersion)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Will install OCP version " + customOCPImage)
exutil.By("config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, customOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("hive.go namespace..." + oc.Namespace())
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
e2e.Logf("test OCP-25310")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("test OCP-33374")
ocpVersion = majorVersion + "." + minorVersion + "." + patchVersion
if ocpVersion == "" {
g.Fail("Case failed because no OCP version extracted from Image")
}
if ocpVersion != "" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, ocpVersion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels}"}).check(oc)
}
e2e.Logf("test OCP-39747")
if ocpVersion != "" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, ocpVersion, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installVersion}"}).check(oc)
}
exutil.By("OCP-23165:Hive supports remote Machine Set Management for AWS")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check worker machinepool .status.replicas = 3")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check only 1 machineset up")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check only one machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 3")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 3}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale up to 3")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 3 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("Patch infra machinepool .spec.replicas to 2")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"replicas": 2}}`}).check(oc)
machinesetsname = getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machinesets scale down to 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1 1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check 2 machines in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
e2e.Logf("OCP-22760: Use custom cluster image set to deploy cluster")
fullImgString := customOCPImage[strings.Index(customOCPImage, ":")+1:]
newCheck("expect", "get", asAdmin, withoutNamespace, compare, fullImgString, ok, DefaultTimeout, []string{"ClusterVersion", "version", "-o=jsonpath={.status.desired.version}", "--kubeconfig=" + kubeconfig}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | e3e0fccd-d4b1-4f15-a30a-02bb099bbbf4 | Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-44945-Low-37528-Low-37527-[HiveSpec] Hive supports ClusterPool runningCount and hibernateAfter[Serial] | ['"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-44945-Low-37528-Low-37527-[HiveSpec] Hive supports ClusterPool runningCount and hibernateAfter[Serial]", func() {
testCaseID := "44945"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
e2e.Logf("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet"}).check(oc)
//secrets can be accessed by pod in the same namespace, so copy pull-secret and aws-creds to target namespace for the pool
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 0,
maxConcurrent: 2,
hibernateAfter: "10m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
e2e.Logf("Check if ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 2, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, FakeClusterInstallTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
e2e.Logf("OCP-44945, step 2: check all cluster are in Hibernating status")
cdListStr := getCDlistfromPool(oc, poolName)
var cdArray []string
cdArray = strings.Split(strings.TrimSpace(cdListStr), "\n")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
e2e.Logf("OCP-37528, step 3: check hibernateAfter and powerState fields")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.spec.powerState}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10m", ok, DefaultTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.spec.hibernateAfter}"}).check(oc)
}
exutil.By("OCP-44945, step 5: Patch .spec.runningCount=1...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"runningCount":1}}`}).check(oc)
e2e.Logf("OCP-44945, step 6: Check the unclaimed clusters in the pool, CD whose creationTimestamp is the oldest becomes Running")
var oldestCD, oldestCDTimestamp string
oldestCDTimestamp = ""
for i := range cdArray {
creationTimestamp := getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdArray[i], "-n", cdArray[i], "-o=jsonpath={.metadata.creationTimestamp}")
e2e.Logf("CD %d is %s, creationTimestamp is %s", i, cdArray[i], creationTimestamp)
if strings.Compare(oldestCDTimestamp, "") == 0 || strings.Compare(oldestCDTimestamp, creationTimestamp) > 0 {
oldestCDTimestamp = creationTimestamp
oldestCD = cdArray[i]
}
}
e2e.Logf("The CD with the oldest creationTimestamp is %s", oldestCD)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD}).check(oc)
exutil.By("OCP-44945, step 7: Patch pool.spec.runningCount=3...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"runningCount":3}}`}).check(oc)
e2e.Logf("OCP-44945, step 7: check runningCount=3 but pool size is still 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.spec.runningCount}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.spec.size}"}).check(oc)
e2e.Logf("OCP-44945, step 7: All CDs in the pool become Running")
for i := range cdArray {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdArray[i], "-n", cdArray[i]}).check(oc)
}
exutil.By("OCP-44945, step 8: Claim a CD from the pool...")
claimTemp := filepath.Join(testDataDir, "clusterclaim.yaml")
claimName := poolName + "-claim"
claim := clusterClaim{
name: claimName,
namespace: oc.Namespace(),
clusterPoolName: poolName,
template: claimTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterClaim", oc.Namespace(), claimName})
claim.create(oc)
e2e.Logf("OCP-44945, step 8: Check the claimed CD is the one whose creationTimestamp is the oldest")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, oldestCD, ok, ClusterResumeTimeout, []string{"ClusterClaim", claimName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("OCP-44945, step 9: Check CD's ClaimedTimestamp is set")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "claimedTimestamp", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.clusterPoolRef}"}).check(oc)
e2e.Logf("OCP-37528, step 5: Check the claimed CD is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("OCP-37528, step 6: Check the claimed CD is in Hibernating status due to hibernateAfter=10m")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout+5*DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
exutil.By("OCP-37527, step 4: patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("OCP-37527, step 5: CD becomes Hibernating again due to hibernateAfter=10m")
//patch makes CD to be Running soon but it needs more time to get back from Hibernation actually so overall timer is ClusterResumeTimeout + hibernateAfter
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout+5*DefaultTimeout, []string{"ClusterDeployment", oldestCD, "-n", oldestCD, "-o=jsonpath={.spec.powerState}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 1d107e1f-4fd0-48d2-9867-af2129d8adf4 | Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-23040-Medium-42113-High-34719-Low-41250-High-25334-High-23876-[HiveSpec] Hive to create SyncSet resource[Serial] | ['"encoding/json"', '"fmt"', '"os"', '"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:jshu-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-23040-Medium-42113-High-34719-Low-41250-High-25334-High-23876-[HiveSpec] Hive to create SyncSet resource[Serial]", func() {
testCaseID := "23040"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create SyncSet for resource apply......")
syncSetName := testCaseID + "-syncset1"
configMapName := testCaseID + "-configmap1"
configMapNamespace := testCaseID + "-" + getRandomString() + "-hive1"
resourceMode := "Sync"
syncTemp := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource := syncSetResource{
name: syncSetName,
namespace: oc.Namespace(),
namespace2: configMapNamespace,
cdrefname: cdName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: resourceMode,
template: syncTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName})
syncResource.create(oc)
e2e.Logf("Check ClusterDeployment is installed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Check if syncSet is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName, ok, DefaultTimeout, []string{"SyncSet", syncSetName, "-n", oc.Namespace()}).check(oc)
exutil.By("Test Syncset Resource part......")
e2e.Logf("OCP-34719, step 3: Check if clustersync and clustersynclease are created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterSyncLease", cdName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("OCP-42113: Check if there is STATUS in clustersync tabular output.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "STATUS", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "MESSAGE", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o", "wide"}).check(oc)
e2e.Logf("OCP-34719, step 4: Check clustersync will record all syncsets first success time.")
successMessage := "All SyncSets and SelectorSyncSets have been applied to the cluster"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, successMessage, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Success", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].result}", syncSetName)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"}).check(oc)
e2e.Logf("OCP-34719, step 5: Check firstSuccessTime won't be changed when there are new syncset created.")
firstSuccessTime, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"))
o.Expect(err).NotTo(o.HaveOccurred())
syncSetName2 := testCaseID + "-syncset2"
configMapName2 := testCaseID + "-configmap2"
configMapNamespace2 := testCaseID + "-" + getRandomString() + "-hive2"
syncTemp2 := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource2 := syncSetResource{
name: syncSetName2,
namespace: oc.Namespace(),
namespace2: configMapNamespace2,
cdrefname: cdName,
ramode: resourceMode,
cmname: configMapName2,
cmnamespace: configMapNamespace2,
template: syncTemp2,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName2})
syncResource2.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName2, ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Success", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].result}", syncSetName2)}).check(oc)
updatedFirstSuccessTime, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterSync", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.firstSuccessTime}"))
o.Expect(err).NotTo(o.HaveOccurred())
if !updatedFirstSuccessTime.Equal(firstSuccessTime) {
e2e.Failf("firstSuccessTime changed when new SyncSet is created")
}
e2e.Logf("Check if configMaps are stored in resourcesToDelete field in ClusterSync CR and they are applied on the target cluster.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapName, "-n", configMapNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapName2, "-n", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName2)}).check(oc)
e2e.Logf("OCP-34719, step 6: Check Resource can be deleted from target cluster via SyncSet when resourceApplyMode is Sync.")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"resourceApplyMode": "Sync"}}`}).check(oc)
patchYaml := `
spec:
resources:
- apiVersion: v1
kind: Namespace
metadata:
name: ` + configMapNamespace2
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
e2e.Logf("Check if ConfigMap %s has deleted from target cluster and clusterSync CR.", configMapName2)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", "-n", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName2, nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"ConfigMap\")].name}", syncSetName2)}).check(oc)
e2e.Logf("OCP-41250: Check Resource won't be deleted from target cluster via SyncSet when resourceApplyMode is Upsert.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Namespace", configMapNamespace2}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"Namespace\")].name}", syncSetName2)}).check(oc)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"resourceApplyMode": "Upsert"}}`}).check(oc)
e2e.Logf("Check if resourcesToDelete field is gone in ClusterSync CR.")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete}", syncSetName2)}).check(oc)
e2e.Logf("Delete Namespace CR from SyncSet, check if Namespace is still exit in target cluster")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetName2, "-n", oc.Namespace(), "--type", "json", "-p", `[{"op": "replace", "path": "/spec/resources", "value":[]}]`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, nok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.status.syncSets[?(@.name==\"%s\")].resourcesToDelete[?(.kind==\"Namespace\")].name}", syncSetName2)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNamespace2, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Namespace", configMapNamespace2}).check(oc)
e2e.Logf("OCP-34719, step 8: Create a bad SyncSet, check if there will be error message in ClusterSync CR.")
syncSetName3 := testCaseID + "-syncset3"
configMapName3 := testCaseID + "-configmap3"
configMapNamespace3 := testCaseID + "-" + getRandomString() + "-hive3"
syncTemp3 := filepath.Join(testDataDir, "syncset-resource.yaml")
syncResource3 := syncSetResource{
name: syncSetName3,
namespace: oc.Namespace(),
namespace2: configMapNamespace3,
cdrefname: cdName,
ramode: resourceMode,
cmname: configMapName3,
cmnamespace: "namespace-non-exist",
template: syncTemp3,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetName3})
syncResource3.create(oc)
errorMessage := fmt.Sprintf("SyncSet %s is failing", syncSetName3)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetName3, ok, DefaultTimeout, []string{"SyncSet", syncSetName3, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, errorMessage, ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Failed")].message}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterSync", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Failed")].status}`}).check(oc)
exutil.By("OCP-23876: Test Syncset Patch part......")
e2e.Logf("Create a test ConfigMap CR on target cluster.")
configMapNameInRemote := testCaseID + "-patch-test"
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "configmap", configMapNameInRemote, "--from-literal=foo=bar", "-n", configMapNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapNameInRemote, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "bar", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
syncSetPatchName := testCaseID + "-syncset-patch"
syncPatchTemp := filepath.Join(testDataDir, "syncset-patch.yaml")
patchContent := `{ "data": { "foo": "baz-strategic" } }`
patchType := "strategic"
syncPatch := syncSetPatch{
name: syncSetPatchName,
namespace: oc.Namespace(),
cdrefname: cdName,
cmname: configMapNameInRemote,
cmnamespace: configMapNamespace,
pcontent: patchContent,
patchType: patchType,
template: syncPatchTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetPatchName})
syncPatch.create(oc)
e2e.Logf("Check if SyncSetPatch is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetPatchName, ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in strategic patch type.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "strategic", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-strategic", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in merge patch type.")
patchYaml = `
spec:
patches:
- apiVersion: v1
kind: ConfigMap
name: ` + configMapNameInRemote + `
namespace: ` + configMapNamespace + `
patch: |-
{ "data": { "foo": "baz-merge" } }
patchType: merge`
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "merge", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-merge", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
e2e.Logf("Check if SyncSetPatch works well when in json patch type.")
patchYaml = `
spec:
patches:
- apiVersion: v1
kind: ConfigMap
name: ` + configMapNameInRemote + `
namespace: ` + configMapNamespace + `
patch: |-
[ { "op": "replace", "path": "/data/foo", "value": "baz-json" } ]
patchType: json`
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "json", ok, DefaultTimeout, []string{"SyncSet", syncSetPatchName, "-n", oc.Namespace(), fmt.Sprintf("-o=jsonpath={.spec.patches[?(@.name==\"%s\")].patchType}", configMapNameInRemote)}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "baz-json", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ConfigMap", configMapNameInRemote, "-n", configMapNamespace, "-o=jsonpath={.data.foo}"}).check(oc)
exutil.By("OCP-25334: Test Syncset SecretReference part......")
syncSetSecretName := testCaseID + "-syncset-secret"
syncSecretTemp := filepath.Join(testDataDir, "syncset-secret.yaml")
sourceName := testCaseID + "-secret"
e2e.Logf("Create temp Secret in current namespace.")
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), sourceName})
err = oc.Run("create").Args("secret", "generic", sourceName, "--from-literal=testkey=testvalue", "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, ok, DefaultTimeout, []string{"Secret", sourceName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check Secret won't exit on target cluster before syncset-secret created.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, nok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Secret", "-n", configMapNamespace}).check(oc)
syncSecret := syncSetSecret{
name: syncSetSecretName,
namespace: oc.Namespace(),
cdrefname: cdName,
sname: sourceName,
snamespace: oc.Namespace(),
tname: sourceName,
tnamespace: configMapNamespace,
template: syncSecretTemp,
}
defer cleanupObjects(oc, objectTableRef{"SyncSet", oc.Namespace(), syncSetSecretName})
syncSecret.create(oc)
e2e.Logf("Check if syncset-secret is created successfully.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, syncSetSecretName, ok, DefaultTimeout, []string{"SyncSet", syncSetSecretName, "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check if the Secret is copied to the target cluster.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, sourceName, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Secret", sourceName, "-n", configMapNamespace}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | eb1c9a8b-7ecb-4e76-a88d-2e9d6a75a441 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-35990-[HiveSpec] Hive support limiting install attempt[Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:jshu-Medium-35990-[HiveSpec] Hive support limiting install attempt[Serial]", func() {
testCaseID := "35990"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
oc.SetupProject()
e2e.Logf("Don't copy AWS platform credentials to make install failed.")
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterDeployment with installAttemptsLimit=0...")
clusterTemp := filepath.Join(testDataDir, "clusterdeployment.yaml")
clusterLimit0 := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 0,
template: clusterTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterLimit0.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "InstallAttemptsLimitReached", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type==\"ProvisionStopped\")].reason}"}).check(oc)
o.Expect(checkResourceNumber(oc, cdName, []string{"pods", "-A"})).To(o.Equal(0))
exutil.By("Delete the ClusterDeployment and recreate it with installAttemptsLimit=1...")
cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterLimit1 := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: clusterTemp,
}
clusterLimit1.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "InstallAttemptsLimitReached", nok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type==\"ProvisionStopped\")].reason}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"pods", "-n", oc.Namespace()}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 9d8c24b9-af3f-4912-9574-e1a6cc9ec9b5 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-63862-Medium-31931-[HiveSpec] MachinePool Supports Public Subnets[Serial] | ['"context"', '"os"', '"path/filepath"', '"strconv"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/service/cloudformation"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"github.com/openshift/openshift-tests-private/test/extended/testdata"', '"k8s.io/apimachinery/pkg/util/sets"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-63862-Medium-31931-[HiveSpec] MachinePool Supports Public Subnets[Serial]", func() {
// Describes a testing scenario
// azs: azs to put in the MachinePool's manifest
// subnets: subnets to put in the MachinePool's manifest
// expectedSubnets: subnets where we expect to find VM(s)
// expectedStatus: status of the InvalidSubnets condition
type scenario struct {
description string
azs []string
subnets []string
expectedSubnets []string
expectedStatus string
}
// Settings
var (
testCaseID = "63862"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
installConfigSecretName = cdName + "-install-config"
clusterImageSetName = cdName + "-imageset"
machinePoolNamePrefix = "infra"
machinePoolReplicas = 2
machinePoolCount = 0
stackName = "vpc-stack-" + testCaseID
azCount = 3
cidr = "10.0.0.0/16"
azsForTesting = []string{AWSRegion + "a", AWSRegion + "b"}
tmpDir = "/tmp/" + testCaseID + "-" + getRandomString()
)
// AWS Clients
var (
cfg = getAWSConfig(oc, AWSRegion)
cloudFormationClient = cloudformation.NewFromConfig(cfg)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
getSubnetByAz = func(subnetIds []string) map[string]string {
describeSubnetsOutput, err := ec2Client.DescribeSubnets(context.Background(), &ec2.DescribeSubnetsInput{
SubnetIds: subnetIds,
})
o.Expect(err).NotTo(o.HaveOccurred())
subnetByAz := make(map[string]string)
for _, subnet := range describeSubnetsOutput.Subnets {
az := aws.ToString(subnet.AvailabilityZone)
subnetId := aws.ToString(subnet.SubnetId)
if existingSubnet, ok := subnetByAz[az]; ok {
e2e.Failf("Subnet %v already exists in AZ %v !", existingSubnet, az)
}
subnetByAz[az] = subnetId
}
return subnetByAz
}
// Returns MachinePool.spec.name
getMachinePoolSpecName = func() string {
return machinePoolNamePrefix + "-" + strconv.Itoa(machinePoolCount)
}
// Returns MachinePool.metadata.name
getMachinePoolFullName = func() string {
return cdName + "-" + getMachinePoolSpecName()
}
createMachinePoolWithSubnetsAndAzs = func(subnets []string, azs []string) {
machinePoolName := getMachinePoolSpecName()
machinePoolManifest := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + "-" + machinePoolName + `
namespace: ` + oc.Namespace() + `
annotations:
# OCP-50051: Day 0 MachineSet Security Group Filters workaround
# We need this tag here as we will be creating VMs outside of the cluster's VPC
hive.openshift.io/extra-worker-security-group: default
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: ` + machinePoolName + `
replicas: ` + strconv.Itoa(machinePoolReplicas) + `
platform:
aws:
rootVolume:
size: 22
type: gp3
type: m5.xlarge`
if len(subnets) > 0 {
machinePoolManifest += "\n subnets:\n - " + strings.Join(subnets, "\n - ")
}
if len(azs) > 0 {
machinePoolManifest += "\n zones:\n - " + strings.Join(azs, "\n - ")
}
machinePoolYamlFile := tmpDir + "/" + machinePoolNamePrefix + ".yaml"
err := os.WriteFile(machinePoolYamlFile, []byte(machinePoolManifest), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
// No need to defer a deletion -- MachinePool VMs are deprovisioned along with the cluster
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", machinePoolYamlFile).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
}
getSubnetsForAzs = func(subnetsByAz map[string]string, azs []string) []string {
var subnets []string
for _, az := range azs {
switch subnet, ok := subnetsByAz[az]; ok {
case true:
subnets = append(subnets, subnet)
default:
e2e.Failf("For %v, no subnet found in AZ %v", subnetsByAz, az)
}
}
return subnets
}
checkMachinePoolStatus = func(target string) bool {
InvalidSubnetsCond :=
getCondition(oc, "MachinePool", getMachinePoolFullName(), oc.Namespace(), "InvalidSubnets")
if status, ok := InvalidSubnetsCond["status"]; !ok || status != target {
e2e.Logf("InvalidSubnets condition %v does not match target status %v, keep polling",
InvalidSubnetsCond, target)
return false
}
return true
}
checkMachinePoolVMsSubnets = func(kubeconfig string, expectedSubnets sets.Set[string]) bool {
// Make sure all instances are created
machinePoolInstancesIds := getMachinePoolInstancesIds(oc, getMachinePoolSpecName(), kubeconfig)
if len(machinePoolInstancesIds) != expectedSubnets.Len() {
e2e.Logf("%v MachinePool instances found (%v expected), keep polling",
len(machinePoolInstancesIds), len(expectedSubnets))
return false
}
// Make sure there's an instance in each expected subnet
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
InstanceIds: machinePoolInstancesIds,
})
o.Expect(err).NotTo(o.HaveOccurred())
for _, reservation := range describeInstancesOutput.Reservations {
instances := reservation.Instances
o.Expect(len(instances)).To(o.Equal(1))
instance := instances[0]
instanceId := aws.ToString(instance.InstanceId)
subnet := aws.ToString(instance.SubnetId)
o.Expect(expectedSubnets.Has(subnet)).To(o.BeTrue())
expectedSubnets.Delete(subnet)
e2e.Logf("Instance %v found in subnet %v", instanceId, subnet)
}
if expectedSubnets.Len() != 0 {
e2e.Logf("Expected subnets without VMs: %v", expectedSubnets.UnsortedList())
}
o.Expect(expectedSubnets.Len()).To(o.Equal(0))
return true
}
)
exutil.By("Creating Cluster")
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, clusterImageSetName, oc.Namespace(), installConfigSecretName, cdName)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Standing up a VPC and subnets which span multiple AZs")
endpointVpcTemp, err :=
testdata.Asset("test/extended/testdata/cluster_operator/hive/cloudformation-endpointvpc-temp.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
e2e.Logf("Deleting CloudFormation stack")
_, err := cloudFormationClient.DeleteStack(context.Background(), &cloudformation.DeleteStackInput{
StackName: aws.String(stackName),
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
e2e.Logf("Creating CloudFormation stack")
_, err = cloudFormationClient.CreateStack(context.Background(), &cloudformation.CreateStackInput{
StackName: aws.String(stackName),
TemplateBody: aws.String(string(endpointVpcTemp)),
Parameters: []cloudFormationTypes.Parameter{
{
ParameterKey: aws.String("AvailabilityZoneCount"),
ParameterValue: aws.String(strconv.Itoa(azCount)),
},
{
ParameterKey: aws.String("VpcCidr"),
ParameterValue: aws.String(cidr),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Making sure the CloudFormation stack is ready")
var vpcId, privateSubnetIds, publicSubnetIds string
waitUntilStackIsReady := func() bool {
describeStackOutput, err := cloudFormationClient.DescribeStacks(context.Background(),
&cloudformation.DescribeStacksInput{
StackName: aws.String(stackName),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeStackOutput.Stacks)).To(o.Equal(1))
stackStatus := describeStackOutput.Stacks[0].StackStatus
if stackStatus != cloudFormationTypes.StackStatusCreateComplete {
e2e.Logf("Stack status = %s, keep polling", stackStatus)
return false
}
// Get stack info once it is ready
for _, output := range describeStackOutput.Stacks[0].Outputs {
switch aws.ToString(output.OutputKey) {
case "VpcId":
vpcId = aws.ToString(output.OutputValue)
case "PrivateSubnetIds":
privateSubnetIds = aws.ToString(output.OutputValue)
case "PublicSubnetIds":
publicSubnetIds = aws.ToString(output.OutputValue)
}
}
return true
}
o.Eventually(waitUntilStackIsReady).WithTimeout(15 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
e2e.Logf("Found VpcId = %s, PrivateSubnetIds = %s, PublicSubnetIds = %s", vpcId, privateSubnetIds, publicSubnetIds)
e2e.Logf("Getting private/public subnets by AZ")
privateSubnetByAz := getSubnetByAz(strings.Split(privateSubnetIds, ","))
publicSubnetByAz := getSubnetByAz(strings.Split(publicSubnetIds, ","))
e2e.Logf("Public subnet by AZ = %v\nPrivate subnet by AZ = %v", publicSubnetByAz, privateSubnetByAz)
// We need to tag the default SG with key=Name since it does not come with a Name tag.
// This name will be used later in the hive.openshift.io/extra-worker-security-group: <sg-name>
// annotation on our MachinePool.
exutil.By("Tagging the default SG of the newly-created VPC")
describeSecurityGroupsOutput, err := ec2Client.DescribeSecurityGroups(context.Background(), &ec2.DescribeSecurityGroupsInput{
Filters: []types.Filter{
{
Name: aws.String("vpc-id"),
Values: []string{vpcId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
// According to our CloudFormation template, there should be one and only one SG -- the default SG
o.Expect(len(describeSecurityGroupsOutput.SecurityGroups)).To(o.Equal(1))
defaultSgId := aws.ToString(describeSecurityGroupsOutput.SecurityGroups[0].GroupId)
e2e.Logf("Found default SG = %v", defaultSgId)
_, err = ec2Client.CreateTags(context.Background(), &ec2.CreateTagsInput{
Resources: []string{defaultSgId},
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("default"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Default SG tagged")
exutil.By("Creating temporary directory")
defer func() {
_ = os.RemoveAll(tmpDir)
}()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Waiting for the CD to be installed")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Getting kubeconfig of the target cluster")
kubeconfig := getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
// For OCP-63862, we only test a number of allowed scenarios here.
exutil.By("Testing the following scenarios: ")
publicSubnetsForTesting := getSubnetsForAzs(publicSubnetByAz, azsForTesting)
privateSubnetsForTesting := getSubnetsForAzs(privateSubnetByAz, azsForTesting)
// Avoid modifying the first argument of append
blendedSubnetsForTesting := append([]string{publicSubnetsForTesting[0]}, privateSubnetsForTesting[1:]...)
allSubnetsForTesting := append([]string{}, publicSubnetsForTesting...)
allSubnetsForTesting = append(allSubnetsForTesting, privateSubnetsForTesting...)
scenarios := []scenario{
{
description: "MachinePool %v is created with some AZs & num-of-az public subnets. " +
"There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: publicSubnetsForTesting,
expectedSubnets: publicSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs & num-of-az private subnets. " +
"There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: privateSubnetsForTesting,
expectedSubnets: privateSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs & a combination of num-of-az " +
"public/private subnets. There is a one-to-one relationship between the AZs and the subnets.",
azs: azsForTesting,
subnets: blendedSubnetsForTesting,
expectedSubnets: blendedSubnetsForTesting,
expectedStatus: "False",
},
{
description: "MachinePool %v is created with some AZs and num-of-az public + num-of-az private subnets. " +
"There is a one-to-one relationship between the AZs and the public subnets. " +
"There is a one-to-one relationship between the AZs and the private subnets.",
azs: azsForTesting,
subnets: allSubnetsForTesting,
expectedSubnets: privateSubnetsForTesting,
expectedStatus: "False",
},
}
for _, scenario := range scenarios {
machinePoolCount++
e2e.Logf(scenario.description, getMachinePoolSpecName())
createMachinePoolWithSubnetsAndAzs(scenario.subnets, scenario.azs)
// Poll until the status of the InvalidSubnets condition is no longer unknown
newCheck("expect", "get", asAdmin, requireNS, compare, "Unknown", nok,
DefaultTimeout, []string{"MachinePool", getMachinePoolFullName(),
`-o=jsonpath={.status.conditions[?(@.type=="InvalidSubnets")].status}`}).check(oc)
// Make sure the status of the InvalidSubnets condition stays expected for a while
o.Consistently(checkMachinePoolStatus).
WithTimeout(2 * time.Minute).
WithPolling(30 * time.Second).
WithArguments(scenario.expectedStatus).
Should(o.BeTrue())
// Make sure VMs are created in expected subnets
o.Eventually(checkMachinePoolVMsSubnets).
WithTimeout(4*time.Minute).
WithPolling(30*time.Second).
WithArguments(kubeconfig, sets.New[string](scenario.expectedSubnets...)).
Should(o.BeTrue())
}
}) | |||||
test case | openshift/openshift-tests-private | 83fd6792-bea8-436d-a375-7cf98080d09b | Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-41212-High-43751-Medium-57403-[HiveSDRosa] [AWSGov] Hive supports to install private cluster [Disruptive] | ['"context"', '"encoding/json"', '"fmt"', '"os"', '"os/exec"', '"path/filepath"', '"reflect"', '"sort"', '"strconv"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"', '"github.com/aws/aws-sdk-go-v2/service/cloudformation"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"github.com/aws/aws-sdk-go-v2/service/iam"', '"github.com/openshift/openshift-tests-private/test/extended/testdata"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-41212-High-43751-Medium-57403-[HiveSDRosa] [AWSGov] Hive supports to install private cluster [Disruptive]", func() {
// Settings
var (
testCaseID = "41212"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
stackName = "endpointvpc-stack-" + testCaseID
// Should not overlap with the CIDR of the associate VPC
cidr = "10.1.0.0/16"
// Number of AZs for the endpoint VPC. Can be different from the equivalent for the associated VPC.
azCount = 3
dnsRecordType = "Alias"
)
// For OCP-43751
if isGovCloud {
dnsRecordType = "ARecord"
}
// AWS Clients
var (
cfg = getAWSConfig(oc, region)
cloudFormationClient = cloudformation.NewFromConfig(cfg)
ec2Client = ec2.NewFromConfig(cfg)
)
// Functions
var (
callCmd = func(cmd *exec.Cmd) error {
e2e.Logf("Calling command: %v", cmd)
out, err := cmd.CombinedOutput()
e2e.Logf("Command output: %s", out)
return err
}
waitForHiveadmissionRedeployment = func(initialHiveConfigGenInt int) bool {
// Make sure HiveConfig's generation is new
hiveConfigGen, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.metadata.generation}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
hiveConfigGenInt, err := strconv.Atoi(hiveConfigGen)
o.Expect(err).NotTo(o.HaveOccurred())
if hiveConfigGenInt <= initialHiveConfigGenInt {
e2e.Logf("HiveConfig generation (%v) <= initial HiveConfig generation (%v), keep polling",
hiveConfigGenInt, initialHiveConfigGenInt)
return false
}
// Make sure the generation is observed
hiveConfigGenObs, _, err := oc.
AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.status.observedGeneration}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
hiveConfigGenObsInt, err := strconv.Atoi(hiveConfigGenObs)
o.Expect(err).NotTo(o.HaveOccurred())
if hiveConfigGenObsInt != hiveConfigGenInt {
e2e.Logf("HiveConfig observed generation (%v) != HiveConfig generation (%v), keep polling",
hiveConfigGenObsInt, hiveConfigGenInt)
return false
}
return true
}
checkCDConditions = func() bool {
awsPrivateLinkFailedCondition :=
getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "AWSPrivateLinkFailed")
if status, ok := awsPrivateLinkFailedCondition["status"]; !ok || status != "False" {
e2e.Logf("For condition AWSPrivateLinkFailed, status = %s, keep polling", status)
return false
}
awsPrivateLinkReadyCondition :=
getCondition(oc, "ClusterDeployment", cdName, oc.Namespace(), "AWSPrivateLinkReady")
if status, ok := awsPrivateLinkReadyCondition["status"]; !ok || status != "True" {
e2e.Logf("For condition AWSPrivateLinkReady, status = %s, keep polling", status)
return false
}
return true
}
compareLocalAndRemoteAllowedPrincipals = func(vpceId string, localAllowedPrincipals []string) bool {
describeVpcEndpointServicePermissionsOutput, err := ec2Client.DescribeVpcEndpointServicePermissions(
context.Background(),
&ec2.DescribeVpcEndpointServicePermissionsInput{
ServiceId: aws.String(vpceId),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
var remoteAllowedPrincipals []string
for _, principal := range describeVpcEndpointServicePermissionsOutput.AllowedPrincipals {
remoteAllowedPrincipals = append(remoteAllowedPrincipals, *principal.Principal)
}
sort.Strings(remoteAllowedPrincipals)
sort.Strings(localAllowedPrincipals)
e2e.Logf("Local allowed principals = %v; remote allowed principals = %v",
localAllowedPrincipals, remoteAllowedPrincipals)
return reflect.DeepEqual(localAllowedPrincipals, remoteAllowedPrincipals)
}
)
exutil.By("Extracting Hiveutil")
tmpDir := "/tmp/" + testCaseID + "-" + getRandomString()
defer func(tempdir string) {
_ = os.RemoveAll(tempdir)
}(tmpDir)
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
hiveutilPath := extractHiveutil(oc, tmpDir)
e2e.Logf("hiveutil extracted to %v", hiveutilPath)
exutil.By("Standing up an endpoint VPC and related resources")
endpointVpcTemp, err :=
testdata.Asset("test/extended/testdata/cluster_operator/hive/cloudformation-endpointvpc-temp.yaml")
o.Expect(err).NotTo(o.HaveOccurred())
defer func() {
// No need to wait for stack deletion to finish. This will save us a couple of minutes.
e2e.Logf("Deleting CloudFormation stack")
_, err := cloudFormationClient.DeleteStack(context.Background(), &cloudformation.DeleteStackInput{
StackName: aws.String(stackName),
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
e2e.Logf("Creating CloudFormation stack")
_, err = cloudFormationClient.CreateStack(context.Background(), &cloudformation.CreateStackInput{
StackName: aws.String(stackName),
TemplateBody: aws.String(string(endpointVpcTemp)),
Parameters: []cloudFormationTypes.Parameter{
{
ParameterKey: aws.String("AvailabilityZoneCount"),
ParameterValue: aws.String(strconv.Itoa(azCount)),
},
{
ParameterKey: aws.String("VpcCidr"),
ParameterValue: aws.String(cidr),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Making sure the CloudFormation stack is ready")
var vpcId, privateSubnetIds string
waitUntilStackIsReady := func() bool {
describeStackOutput, err := cloudFormationClient.DescribeStacks(context.Background(),
&cloudformation.DescribeStacksInput{
StackName: aws.String(stackName),
},
)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeStackOutput.Stacks)).To(o.Equal(1))
stackStatus := describeStackOutput.Stacks[0].StackStatus
if stackStatus != cloudFormationTypes.StackStatusCreateComplete {
e2e.Logf("Stack status = %s, keep polling", stackStatus)
return false
}
// Get stack info once it is ready
for _, output := range describeStackOutput.Stacks[0].Outputs {
switch aws.ToString(output.OutputKey) {
case "VpcId":
vpcId = aws.ToString(output.OutputValue)
case "PrivateSubnetIds":
privateSubnetIds = aws.ToString(output.OutputValue)
}
}
return true
}
o.Eventually(waitUntilStackIsReady).WithTimeout(15 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
e2e.Logf("VpcId = %s, PrivateSubnetIds = %s", vpcId, privateSubnetIds)
// For OCP-43751:
// The hiveutil awsprivatelink add/remove commands filter out the private route tables through the Name tag.
// On AWS Gov cloud openshift clusters are installed into a BYO VPC,
// which is often created as part of a CloudFormation stack.
// There is no guarantee that the private route tables,
// which belong to the same CloudFormation stack,
// has a Name (or whatever pre-defined) tag.
// Consequently, we need to tag these route tables ourselves.
// This makes the test cases disruptive.
if isGovCloud {
infraId, err := exutil.GetInfraID(oc)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found infraID = %v", infraId)
// Get resources to tag
describeVpcsOutput, err := ec2Client.DescribeVpcs(context.Background(), &ec2.DescribeVpcsInput{
Filters: []types.Filter{
{
Name: aws.String("tag-key"),
Values: []string{fmt.Sprintf("kubernetes.io/cluster/%v", infraId)},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeVpcsOutput.Vpcs)).To(o.Equal(1))
associatedVpcId := aws.ToString(describeVpcsOutput.Vpcs[0].VpcId)
e2e.Logf("Found associated VPC ID = %v", associatedVpcId)
describeRouteTableOutput, err := ec2Client.DescribeRouteTables(context.Background(), &ec2.DescribeRouteTablesInput{
Filters: []types.Filter{
{
Name: aws.String("vpc-id"),
Values: []string{associatedVpcId},
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
associatedRouteTables := describeRouteTableOutput.RouteTables
o.Expect(len(associatedRouteTables)).NotTo(o.BeZero())
var privateAssociatedRouteTableIds []string
for _, associatedRouteTable := range associatedRouteTables {
associatedRouteTableId := aws.ToString(associatedRouteTable.RouteTableId)
e2e.Logf("Found associated route table %v", associatedRouteTableId)
for _, route := range associatedRouteTable.Routes {
if natGatewayId := aws.ToString(route.NatGatewayId); natGatewayId != "" {
e2e.Logf("Found a route targeting a NAT gateway, route table %v is private", associatedRouteTableId)
privateAssociatedRouteTableIds = append(privateAssociatedRouteTableIds, associatedRouteTableId)
break
}
}
}
o.Expect(len(privateAssociatedRouteTableIds)).NotTo(o.BeZero())
// Tagging
e2e.Logf("Tagging %v with Name = private", privateAssociatedRouteTableIds)
defer func() {
_, err := ec2Client.DeleteTags(context.Background(), &ec2.DeleteTagsInput{
Resources: privateAssociatedRouteTableIds,
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("private"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
}()
_, err = ec2Client.CreateTags(context.Background(), &ec2.CreateTagsInput{
Resources: privateAssociatedRouteTableIds,
Tags: []types.Tag{
{
Key: aws.String("Name"),
Value: aws.String("private"),
},
},
})
o.Expect(err).NotTo(o.HaveOccurred())
}
// Some (idempotent) awsprivatelink subcommands below are polled until succeed.
// Rationale:
// Calling an awsprivatelink subcommand immediately after another might fail
// due to etcd being only eventually consistent (as opposed to strongly consistent).
// In fact, awsprivatelink subcommands often starts off GETTING resources,
// which are processed and UPDATED before the command terminates.
// As a result, the later command might end up getting stale resources,
// causing the UPDATE request it makes to fail.
exutil.By("Setting up privatelink")
defer func() {
cmd := exec.Command(hiveutilPath, "awsprivatelink", "disable", "-d")
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
}()
// This is the first awsprivatelink subcommand, so no need to poll
cmd := exec.Command(
hiveutilPath, "awsprivatelink",
"enable",
"--creds-secret", "kube-system/aws-creds",
"--dns-record-type", dnsRecordType,
"-d")
err = callCmd(cmd)
o.Expect(err).NotTo(o.HaveOccurred())
// Sleep for a few seconds, so the first round of polling is more likely to succeed
time.Sleep(5 * time.Second)
// Get HiveConfig's generation, which will be used to make sure HiveConfig is updated.
initialHiveConfigGen, _, err := oc.AsAdmin().
WithoutNamespace().
Run("get").
Args("hiveconfig/hive", "-o=jsonpath={.metadata.generation}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
initialHiveConfigGenInt, err := strconv.Atoi(initialHiveConfigGen)
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found HiveConfig generation = %v after enabling awsprivatelink", initialHiveConfigGenInt)
e2e.Logf("Setting up endpoint VPC")
defer func() {
cmd := exec.Command(
hiveutilPath, "awsprivatelink",
"endpointvpc", "remove", vpcId,
"--creds-secret", "kube-system/aws-creds",
"-d",
)
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
}()
cmd = exec.Command(
hiveutilPath, "awsprivatelink",
"endpointvpc", "add", vpcId,
"--region", region,
"--creds-secret", "kube-system/aws-creds",
"--subnet-ids", privateSubnetIds,
"-d",
)
o.Eventually(callCmd).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).WithArguments(cmd).Should(o.BeNil())
// It is necessary to wait for the re-deployment of Hive-admission, otherwise the CD gets rejected.
exutil.By("Waiting for the re-deployment of Hive-admission")
o.Eventually(waitForHiveadmissionRedeployment).
WithTimeout(3 * time.Minute).
WithPolling(1 * time.Minute).
WithArguments(initialHiveConfigGenInt).
Should(o.BeTrue())
// Wait until the new hiveadmission Deployment is available
err = oc.
AsAdmin().
WithoutNamespace().
Run("wait").
Args("deploy/hiveadmission", "-n", HiveNamespace, "--for", "condition=available=true", "--timeout=3m").
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Creating ClusterImageSet")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret")
// Reuse endpoint VPC here to save cloud resource.
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfigPrivateLink{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: basedomain,
name2: cdName,
region: region,
publish: PublishInternal,
machineNetworkCidr: cidr,
privateSubnetId1: strings.Split(privateSubnetIds, ",")[0],
privateSubnetId2: strings.Split(privateSubnetIds, ",")[1],
privateSubnetId3: strings.Split(privateSubnetIds, ",")[2],
template: filepath.Join(testDataDir, "aws-install-config-privatelink.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Copying AWS credentials")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copying pull secret")
createPullSecret(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment")
clusterDeployment := clusterDeploymentPrivateLink{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: basedomain,
clusterName: cdName,
credRef: AWSCreds,
region: region,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment-aws-privatelink.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
clusterDeployment.create(oc)
exutil.By("OCP-57403: Support to add AdditionalAllowedPrincipals for PrivateLink VPCE Services")
e2e.Logf("Waiting for awsprivatelink reconcile to be done")
err = oc.
AsAdmin().
Run("wait").
Args("ClusterDeployment", cdName, "--for", "condition=AWSPrivateLinkReady=true", "--timeout=20m").
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Getting default allowed principal")
defaultAllowedPrincipalJsonPath :=
"{.status.platformStatus.aws.privateLink.vpcEndpointService.defaultAllowedPrincipal}"
defaultAllowedPrincipal, _, err := oc.
AsAdmin().
Run("get").
Args("ClusterDeployment", cdName, "-o=jsonpath="+defaultAllowedPrincipalJsonPath).
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(defaultAllowedPrincipal).To(o.HavePrefix(fmt.Sprintf("arn:%v:iam::", awsPartition)))
e2e.Logf("Found defaultAllowedPrincipal = %s", defaultAllowedPrincipal)
e2e.Logf("Getting vpce ID")
vpceIdJsonPath :=
"{.status.platformStatus.aws.privateLink.vpcEndpointService.id}"
vpceId, _, err := oc.
AsAdmin().
Run("get").
Args("ClusterDeployment", cdName, "-o=jsonpath="+vpceIdJsonPath).
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(vpceId).To(o.HavePrefix("vpce-svc-"))
e2e.Logf("Found vpce ID = %s", vpceId)
e2e.Logf("Adding an additionalAllowedPrincipal")
additionalAllowedPrincipal := fmt.Sprintf("arn:%v:iam::301721915996:user/fakefxie", awsPartition)
additionalAllowedPrincipalPatch := `
spec:
platform:
aws:
privateLink:
additionalAllowedPrincipals:
- ` + additionalAllowedPrincipal
err = oc.
AsAdmin().
Run("patch").
Args("ClusterDeployment", cdName, "--type", "merge", "-p", additionalAllowedPrincipalPatch).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Comparing local & remote allowed principals")
localAllowedPrincipals := []string{defaultAllowedPrincipal, additionalAllowedPrincipal}
o.Eventually(compareLocalAndRemoteAllowedPrincipals).
WithTimeout(5*time.Minute).
WithPolling(30*time.Second).
WithArguments(vpceId, localAllowedPrincipals).
Should(o.BeTrue())
e2e.Logf("Removing additionalAllowedPrincipals from CD")
err = oc.
AsAdmin().
Run("patch").
Args("ClusterDeployment", cdName, "--type=json", "-p",
`[{"op":"remove", "path": "/spec/platform/aws/privateLink/additionalAllowedPrincipals"}]`).
Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Comparing local & remote allowed principals again")
localAllowedPrincipals = []string{defaultAllowedPrincipal}
o.Eventually(compareLocalAndRemoteAllowedPrincipals).
WithTimeout(5*time.Minute).
WithPolling(30*time.Second).
WithArguments(vpceId, localAllowedPrincipals).
Should(o.BeTrue())
exutil.By("Back to OCP-41212: Waiting for installation to finish")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Checking CD.status")
o.Eventually(checkCDConditions).WithTimeout(3 * time.Minute).WithPolling(1 * time.Minute).Should(o.BeTrue())
privateLinkStatus, _, err := oc.
AsAdmin().
Run("get").
Args("clusterdeployment", cdName, "-o", "jsonpath={.status.platformStatus.aws.privateLink}").
Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Now CD.status.platformStatus.aws.privateLink looks like: \n%s", privateLinkStatus)
// Open question: should we check if the IDs in privateLinkStatus are correct ?
o.Expect(strings.Contains(privateLinkStatus, "hostedZoneID")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "vpcEndpointID")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "vpcEndpointService")).To(o.BeTrue())
o.Expect(strings.Contains(privateLinkStatus, "defaultAllowedPrincipal")).To(o.BeTrue())
exutil.By("Making sure the private target cluster is not directly reachable")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
_, _, err = oc.AsAdmin().WithoutNamespace().Run("get").Args("co", "--kubeconfig", kubeconfig).Outputs()
o.Expect(err).To(o.HaveOccurred())
exutil.By("Making sure the target cluster is reachable from the Hive cluster")
// Due to the PrivateLink networking setup (through awsprivatelink subcommands called above),
// the target cluster can only be accessed from worker nodes of the Hive cluster.
// This is not a problem for the Hive operator, as its Pods are deployed on the worker nodes by default.
selectors := map[string]string{
"node-role.kubernetes.io/worker": "",
}
workerNodeNames := getNodeNames(oc, selectors)
kubeconfigByteSlice, err := os.ReadFile(kubeconfig)
o.Expect(err).NotTo(o.HaveOccurred())
// Ensure literal interpretation by Bash
kubeconfigSingleQuotedStr := "'" + string(kubeconfigByteSlice) + "'"
// Take care of the SCC setup
output, err := exutil.DebugNode(oc, workerNodeNames[0], "bash", "-c",
fmt.Sprintf("echo %s > kubeconfig; oc get co --kubeconfig kubeconfig", kubeconfigSingleQuotedStr))
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("cloud-credential"))
}) | |||||
test case | openshift/openshift-tests-private | f5b8650e-fcb3-47f6-bb7e-8c0c15596ddc | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:liangli-Medium-32223-Medium-35193-High-23308-Hive ClusterDeployment Check installed and uninstalled [Serial] | ['"fmt"', '"os"', '"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"', '"github.com/openshift/openshift-tests-private/test/extended/util/architecture"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:liangli-Medium-32223-Medium-35193-High-23308-Hive ClusterDeployment Check installed and uninstalled [Serial]", func() {
testCaseID := "32223"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Get multi-arch image required for cross-architecture provisioning")
multiarchImage, err := exutil.GetLatestImage("multi", "ocp", "4-dev-preview-multi")
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
vmType: AWSVmTypeARM64,
arch: archARM64,
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, multiarchImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("test OCP-23308: Hive install log does not contain admin credentials, but contains REDACTED LINE OF OUTPUT")
provisionPodName := getProvisionPodNames(oc, cdName, oc.Namespace())[0]
cmd, stdout, err := oc.Run("logs").Args("-f", provisionPodName, "-c", "hive").BackgroundRC()
defer cmd.Process.Kill()
o.Expect(err).NotTo(o.HaveOccurred())
f := stdout.(*os.File)
defer f.Close()
targetLines := []string{
fmt.Sprintf("Access the OpenShift web-console here: https://console-openshift-console.apps.%v.%v\"", cdName, AWSBaseDomain),
"REDACTED LINE OF OUTPUT",
}
targetFound := assertLogs(f, targetLines, nil, 3*time.Minute)
o.Expect(targetFound).To(o.BeTrue())
exutil.By("test OCP-32223 check install")
provisionName, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.provisionRef.name}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(provisionName).NotTo(o.BeEmpty())
e2e.Logf("test OCP-32223 install")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "true", ok, DefaultTimeout, []string{"job", provisionName + "-provision", "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels.hive\\.openshift\\.io/install}"}).check(oc)
exutil.By("test OCP-35193 check uninstall")
e2e.Logf("get aws_access_key_id by secretName")
awsAccessKeyID, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("secret", "aws-creds", "-n", oc.Namespace(), "-o=jsonpath={.data.aws_access_key_id}").Output()
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(provisionName).NotTo(o.BeEmpty())
e2e.Logf("Modify aws creds to invalid")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "-n", oc.Namespace(), "-p", `{"data":{"aws_access_key_id":null}}`, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("delete ClusterDeployment")
_, _, _, err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", cdName, "-n", oc.Namespace()).Background()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationFailed", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationFailed", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].reason}`}).check(oc)
e2e.Logf("Change aws creds to valid again")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("secret", "aws-creds", "-n", oc.Namespace(), "-p", `{"data":{"aws_access_key_id":"`+awsAccessKeyID+`"}}`, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationSucceeded", ok, DefaultTimeout, []string{"clusterdeprovision", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="AuthenticationFailure")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "AuthenticationSucceeded", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DeprovisionLaunchError")].reason}`}).check(oc)
exutil.By("test OCP-32223 check uninstall")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "true", ok, DefaultTimeout, []string{"job", cdName + "-uninstall", "-n", oc.Namespace(), "-o=jsonpath={.metadata.labels.hive\\.openshift\\.io/uninstall}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | fd48f14d-eb77-4eb6-8556-2c4b760cc959 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-33642-[HiveSpec] Hive supports cluster hibernation [Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-Medium-33642-[HiveSpec] Hive supports cluster hibernation [Serial]", func() {
testCaseID := "33642"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check AWS ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Check CD has Hibernating condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
exutil.By("patch the CD to Hibernating...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Hibernating"}}`}).check(oc)
e2e.Logf("Wait for CD to be Hibernating")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Hibernating", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
exutil.By("patch the CD to Running...")
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "--type", "merge", "-p", `{"spec":{"powerState": "Running"}}`}).check(oc)
e2e.Logf("Wait for CD to be Running")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Running", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.powerState}"}).check(oc)
e2e.Logf("Check cd's condition")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Hibernating")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "True", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Ready")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="Unreachable")].status}`}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | cdcd8773-789b-4205-9311-94f2a89bc2b4 | Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-63275-[HiveSpec] Hive support for AWS IMDSv2 [Serial] | ['"context"', '"fmt"', '"os"', '"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/service/ec2"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:fxie-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-63275-[HiveSpec] Hive support for AWS IMDSv2 [Serial]", func() {
var (
testCaseID = "63275"
cdName = "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
workerMpName = "worker"
infraMpName = "infra"
infraMpName2 = "infra-2"
)
exutil.By("Creating ClusterDeployment")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
clusterDeployment := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanCD(oc, clusterDeployment.name+"-imageset", oc.Namespace(), installConfigSecret.name1, clusterDeployment.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, clusterDeployment)
exutil.By("Wait for the cluster to be installed")
newCheck("expect", "get", asAdmin, requireNS, compare, "true", ok,
ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("Creating temporary directory")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer func() {
_ = os.RemoveAll(tmpDir)
}()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Saving kubeconfig of the target cluster")
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
exutil.By("Creating worker MachinePool with metadataService.authentication un-specified")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: filepath.Join(testDataDir, "machinepool-worker-aws.yaml"),
}
workermp.create(oc)
exutil.By("Creating infra MachinePool with metadataService.authentication = Optional")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
authentication: "Optional",
template: filepath.Join(testDataDir, "machinepool-infra-aws.yaml"),
}
defer cleanupObjects(oc, objectTableRef{
"MachinePool", oc.Namespace(), fmt.Sprintf("%s-%s", cdName, infraMpName),
})
inframp.create(oc)
exutil.By("Creating another infra MachinePool with metadataService.authentication = Required")
fullInframpName2 := fmt.Sprintf("%s-%s", cdName, infraMpName2)
inframp2 := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + fullInframpName2 + `
namespace: ` + oc.Namespace() + `
spec:
clusterDeploymentRef:
name: ` + cdName + `
name: ` + infraMpName2 + `
platform:
aws:
metadataService:
authentication: Required
rootVolume:
size: 22
type: gp2
type: m4.xlarge
replicas: 1`
filename := tmpDir + "/" + testCaseID + infraMpName2
err = os.WriteFile(filename, []byte(inframp2), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), fullInframpName2})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Extracting Instance IDs")
instanceIdByMachinePool := make(map[string]string)
machinePools := []string{workerMpName, infraMpName, infraMpName2}
getInstanceIds := func() bool {
for _, machinePool := range machinePools {
instanceIds := getMachinePoolInstancesIds(oc, machinePool, kubeconfig)
if len(instanceIds) == 0 {
e2e.Logf("%s Machines not found, keep polling", machinePool)
return false
}
instanceIdByMachinePool[machinePool] = instanceIds[0]
}
return true
}
o.Eventually(getInstanceIds).WithTimeout(10 * time.Minute).WithPolling(30 * time.Second).Should(o.BeTrue())
e2e.Logf("Instance IDs found: %v", instanceIdByMachinePool)
exutil.By("Checking IMDSv2 settings")
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
expectedIMDSv2 := map[string]string{
workerMpName: "optional",
infraMpName: "optional",
infraMpName2: "required",
}
for machinePool, instanceId := range instanceIdByMachinePool {
e2e.Logf("Checking IDMSv2 settings on a %s instance", machinePool)
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
InstanceIds: []string{instanceId},
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeInstancesOutput.Reservations)).To(o.Equal(1))
o.Expect(len(describeInstancesOutput.Reservations[0].Instances)).To(o.Equal(1))
o.Expect(string(describeInstancesOutput.Reservations[0].Instances[0].MetadataOptions.HttpTokens)).
To(o.Equal(expectedIMDSv2[machinePool]))
// Limit the frequency of API calls
time.Sleep(5 * time.Second)
}
}) | |||||
test case | openshift/openshift-tests-private | bb0d8627-3c3e-4173-a6c3-d0b74053fe56 | Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-49471-High-76460-High-23677- Change EC2RootVolume: make IOPS optional [Serial] | ['"fmt"', '"os"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Medium-49471-High-76460-High-23677- Change EC2RootVolume: make IOPS optional [Serial]", func() {
testCaseID := "49471"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret with iops=1...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool with IOPS optional ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
iops: 2,
template: workermachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
workermp.create(oc)
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
iops: 1,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
exutil.By("OCP-76460:Creating infra2 MachinePool with machineLabels configuration ......")
var infraMpName2 = "infra2"
tmpDir76460 := "/tmp/" + cdName + "-" + getRandomString()
defer func() {
_ = os.RemoveAll(tmpDir76460)
}()
err := os.MkdirAll(tmpDir76460, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
fullInframpName2 := fmt.Sprintf("%s-%s", cdName, infraMpName2)
inframp2 := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + fullInframpName2 + `
namespace: ` + oc.Namespace() + `
spec:
machineLabels:
mihuang-test-label: mihaung-test-value
machine.openshift.io/cluster-api-cluster: test-conflict
clusterDeploymentRef:
name: ` + cdName + `
name: ` + infraMpName2 + `
platform:
aws:
rootVolume:
size: 120
type: gp3
type: m6a.xlarge
replicas: 1`
filename := tmpDir76460 + "/" + testCaseID + infraMpName2
err = os.WriteFile(filename, []byte(inframp2), 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), fullInframpName2})
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
e2e.Logf("Check worker machinepool .spec.platform.aws.rootVolume.iops = 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.rootVolume.iops}"}).check(oc)
e2e.Logf("Check infra machinepool .spec.platform.aws.rootVolume.iops = 1")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.rootVolume.iops}"}).check(oc)
exutil.By("OCP-23677: Allow modification of machine pool labels and taints")
e2e.Logf("Patching machinepool ...")
patchYaml := `
spec:
taints:
- effect: foo
key: bar
labels:
baz: qux`
err = oc.AsAdmin().Run("patch").Args("MachinePool", cdName+"-worker", "--type", "merge", "-p", patchYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Extracting kubeconfig from remote cluster ...")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
defer os.RemoveAll(tmpDir)
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
infraID, _, err := oc.AsAdmin().Run("get").Args("cd", cdName, "-o", "jsonpath='{.spec.clusterMetadata.infraID}'").Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
infraID = strings.Trim(infraID, "'")
machineSetName := infraID + "-worker-" + AWSRegion + "a"
e2e.Logf("Checking taints & labels on MachineSet %v ...", machineSetName)
expectedTaints := "{\"effect\":\"foo\",\"key\":\"bar\"}"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedTaints, ok, DefaultTimeout, []string{"MachineSet", machineSetName, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.spec.taints[0]}'"}).check(oc)
expectedLabels := "{\"baz\":\"qux\"}"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedLabels, ok, DefaultTimeout, []string{"MachineSet", machineSetName, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.spec.metadata.labels}'"}).check(oc)
exutil.By("OCP-76460: Check the machinepool's status.ownedMachineLabels, only labels that do not conflict with the installer should be marked, while conflicting labels should not be marked.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "mihuang-test-label", ok, DefaultTimeout, []string{"MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "machine.openshift.io/cluster-api-cluster", nok, DefaultTimeout, []string{"MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'"}).check(oc)
exutil.By("OCP-76460: Log in to the spoke cluster and check the machinesets. They should ignore the conflicting labels generated by the installer code and add the new machine labels.")
machineSetName2 := infraID + "-" + infraMpName2 + "-" + AWSRegion + "a"
expectedLabels = "\"mihuang-test-label\":\"mihaung-test-value\""
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedLabels, ok, DefaultTimeout, []string{"MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels}'"}).check(oc)
e2e.Logf("Check the machinesets labels should ignore conflicts generated by the installer code.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, infraID, ok, DefaultTimeout, []string{"MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig=" + kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels.machine\\.openshift\\.io/cluster-api-cluster}'"}).check(oc)
exutil.By("OCP-76460: Update the machinelabels in the machinepool and check that the machinesets ignore the conflicts generated by the installer code.")
e2e.Logf("Patching infra2 machinepool ...")
patchYaml2 := `
spec:
machineLabels:
mihuangtest-conflict-label2: mihuang-test-conflict2
machine.openshift.io/cluster-api-machine-role: test-conflict`
err = oc.AsAdmin().Run("patch").Args("MachinePool", fullInframpName2, "--type", "merge", "-p", patchYaml2).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("OCP-76460: Check that the machinepool's status.ownedMachineLabels updates correctly.")
o.Eventually(func() bool {
out, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("MachinePool", fullInframpName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.ownedMachineLabels}'").Output()
if err != nil {
return false
}
if strings.Contains(out, "mihuangtest-conflict-label2") && !strings.Contains(out, "machine.openshift.io/cluster-api-machine-role") {
return true
}
return false
}, DefaultTimeout, 60*time.Second).Should(o.BeTrue())
e2e.Logf("OCP-76460: Log in to the spoke cluster and check that the machinesets update correctly.")
o.Eventually(func() bool {
out, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("MachineSet", machineSetName2, "-n=openshift-machine-api", "--kubeconfig="+kubeconfig, "-o=jsonpath='{.spec.template.metadata.labels}'").Output()
if err != nil {
return false
}
if strings.Contains(out, "\"mihuangtest-conflict-label2\":\"mihuang-test-conflict2\"") && strings.Contains(out, fmt.Sprintf("\"machine.openshift.io/cluster-api-machine-role\":\"%s\"", infraMpName2)) {
return true
}
return false
}, DefaultTimeout, 60*time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | 99b46628-74fc-4a83-8bf5-cf572f472d20 | Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-24088-Medium-33045-[HiveSpec] Provisioning clusters on AWS with managed dns [Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:mihuang-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-24088-Medium-33045-[HiveSpec] Provisioning clusters on AWS with managed dns [Serial]", func() {
testCaseID := "24088"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Create Route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check Aws ClusterDeployment installed flag is true")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-33045 - Prevent ClusterDeployment deletion until managed DNSZone is gone")
exutil.By("Delete route53-aws-creds in hive namespace")
err := oc.AsAdmin().WithoutNamespace().Run("delete").Args("secret", "route53-aws-creds", "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Try to delete cd")
cmd, _, _, _ := oc.AsAdmin().WithoutNamespace().Run("delete").Args("cd", cdName, "-n", oc.Namespace()).Background()
defer cmd.Process.Kill()
exutil.By("Check the deprovision pod is completed")
DeprovisionPodName := getDeprovisionPodName(oc, cdName, oc.Namespace())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Completed", ok, ClusterUninstallTimeout, []string{"pod", DeprovisionPodName, "-n", oc.Namespace()}).check(oc)
exutil.By("Check the cd is not removed")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace()}).check(oc)
exutil.By("Check the dnszone is not removed")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, ok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
exutil.By("Create route53-aws-creds in hive namespace")
createRoute53AWSCreds(oc, oc.Namespace())
exutil.By("Wait until dnszone controller next reconcile, verify dnszone and cd are removed.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", oc.Namespace()}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, cdName, nok, DefaultTimeout, []string{"dnszone", "-n", oc.Namespace()}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 1853902b-9ae2-4b46-909f-8240a474c89e | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-51195-DNSNotReadyTimeout should be terminal[Disruptive] | ['"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-51195-DNSNotReadyTimeout should be terminal[Disruptive]", func() {
testCaseID := "51195"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Remove Route53-aws-creds in hive namespace if exists to make DNSNotReady")
cleanupObjects(oc, objectTableRef{"secret", HiveNamespace, "route53-aws-creds"})
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: cdName + "." + AWSBaseDomain,
clusterName: cdName,
manageDNS: true,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check DNSNotReady, Provisioned and ProvisionStopped condiitons")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].status}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "DNS Zone not yet available", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].message}`}).check(oc)
e2e.Logf("Check PROVISIONSTATUS=ProvisionStopped ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ProvisionStopped", ok, ClusterResumeTimeout+DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.conditions[?(@.type=='Provisioned')].reason}"}).check(oc)
e2e.Logf("check ProvisionStopped=true and DNSNotReady.reason=DNSNotReadyTimedOut ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "DNSNotReadyTimedOut", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].reason}`}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="ProvisionStopped")].status}`}).check(oc)
exutil.By("Check DNSNotReadyTimeOut beacuse the default timeout is 10 min")
creationTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.metadata.creationTimestamp}"))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get cluster create timestamp,creationTimestampp is %v", creationTimestamp)
dnsNotReadyTimedOuTimestamp, err := time.Parse(time.RFC3339, getResource(oc, asAdmin, withoutNamespace, "ClusterDeployment", cdName, "-n", oc.Namespace(), `-o=jsonpath={.status.conditions[?(@.type=="DNSNotReady")].lastProbeTime}`))
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("get dnsnotready timestap, dnsNotReadyTimedOuTimestamp is %v", dnsNotReadyTimedOuTimestamp)
difference := dnsNotReadyTimedOuTimestamp.Sub(creationTimestamp)
e2e.Logf("default timeout is %v mins", difference.Minutes())
o.Expect(difference.Minutes()).Should(o.BeNumerically(">=", 10))
}) | |||||
test case | openshift/openshift-tests-private | f0dcb37e-7647-4646-9c97-2cb1172b30c5 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-23676-Create cluster with master terminated by manipulation[Serial] | ['"context"', '"path/filepath"', '"strings"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-High-23676-Create cluster with master terminated by manipulation[Serial]", func() {
testCaseID := "23676"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
exutil.By("Creating Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Creating ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Getting infraID from CD...")
var infraID string
var err error
getInfraIDFromCD := func() bool {
infraID, _, err = oc.AsAdmin().Run("get").Args("cd", cdName, "-o=jsonpath={.spec.clusterMetadata.infraID}").Outputs()
return err == nil && strings.HasPrefix(infraID, cdName)
}
o.Eventually(getInfraIDFromCD).WithTimeout(10 * time.Minute).WithPolling(5 * time.Second).Should(o.BeTrue())
e2e.Logf("Found infraID = %v", infraID)
// Get AWS client
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
exutil.By("Waiting until the master VMs are created...")
var describeInstancesOutput *ec2.DescribeInstancesOutput
waitUntilMasterVMCreated := func() bool {
describeInstancesOutput, err = ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag:Name"),
// Globbing leads to filtering AFTER returning a page of instances
// This results in the necessity of looping through pages of instances,
// i.e. some extra complexity.
Values: []string{infraID + "-master-0", infraID + "-master-1", infraID + "-master-2"},
},
},
MaxResults: aws.Int32(6),
})
return err == nil && len(describeInstancesOutput.Reservations) == 3
}
o.Eventually(waitUntilMasterVMCreated).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
// Terminate all master VMs so the Kubernetes API is never up. Provision may fail at earlier stages though.
exutil.By("Terminating the master VMs...")
var instancesToTerminate []string
for _, reservation := range describeInstancesOutput.Reservations {
instancesToTerminate = append(instancesToTerminate, *reservation.Instances[0].InstanceId)
}
_, err = ec2Client.TerminateInstances(context.Background(), &ec2.TerminateInstancesInput{
InstanceIds: instancesToTerminate,
})
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Terminating master VMs %v", instancesToTerminate)
// The stage at which provision fails is not guaranteed. Here we just make sure provision actually fails.
exutil.By("Waiting for the first provision Pod to fail...")
provisionPod1 := getProvisionPodNames(oc, cdName, oc.Namespace())[0]
newCheck("expect", "get", asAdmin, requireNS, compare, "Failed", ok, 1800, []string{"pod", provisionPod1, "-o=jsonpath={.status.phase}"}).check(oc)
exutil.By("Waiting for the second provision Pod to be created...")
waitForProvisionPod2 := func() bool {
provisionPodNames := getProvisionPodNames(oc, cdName, oc.Namespace())
if len(provisionPodNames) > 1 {
e2e.Logf("The second provision Pod is created. Current provision Pods = %v", provisionPodNames)
return true
}
return false
}
o.Eventually(waitForProvisionPod2).WithTimeout(10 * time.Minute).WithPolling(10 * time.Second).Should(o.BeTrue())
}) | |||||
test case | openshift/openshift-tests-private | f58ee3c3-f366-44cc-943c-498003a77c8f | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-23970-The cluster name is limited by 63 characters[Serial] | ['"encoding/json"', '"os"', '"path/filepath"', '"strconv"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-23970-The cluster name is limited by 63 characters[Serial]", func() {
testCaseID := "23970"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Creating ClusterImageSet ...")
clusterImageSetName := cdName + "-imageset"
imageSet := clusterImageSet{
name: clusterImageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName})
imageSet.create(oc)
exutil.By("Creating install-config Secret ...")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Creating pull-secret ...")
createPullSecret(oc, oc.Namespace())
exutil.By("Copying AWS credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Creating ClusterDeployment with a 64-character-long cluster name ...")
clusterName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen] + "-" + "123456789012345678901234567890123456789012345"
clusterDeployment := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: clusterName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
parameters := []string{
"--ignore-unknown-parameters=true",
"-f", clusterDeployment.template,
"-p", "FAKE=" + clusterDeployment.fake,
"NAME=" + clusterDeployment.name,
"NAMESPACE=" + clusterDeployment.namespace,
"BASEDOMAIN=" + clusterDeployment.baseDomain,
"CLUSTERNAME=" + clusterDeployment.clusterName,
"MANAGEDNS=" + strconv.FormatBool(clusterDeployment.manageDNS),
"PLATFORMTYPE=" + clusterDeployment.platformType,
"CREDREF=" + clusterDeployment.credRef,
"REGION=" + clusterDeployment.region,
"IMAGESETREF=" + clusterDeployment.imageSetRef,
"INSTALLCONFIGSECRET=" + clusterDeployment.installConfigSecret,
"PULLSECRETREF=" + clusterDeployment.pullSecretRef,
"INSTALLATTEMPTSLIMIT=" + strconv.Itoa(clusterDeployment.installAttemptsLimit),
}
// Manually create CD to capture the output of oc apply -f cd_manifest_file
var cfgFileJSON string
defer func() {
if err := os.RemoveAll(cfgFileJSON); err != nil {
e2e.Logf("Error removing file %v: %v", cfgFileJSON, err.Error())
}
}()
cfgFileJSON, err := oc.AsAdmin().Run("process").Args(parameters...).OutputToFile(getRandomString() + "-hive-resource-cfg.json")
o.Expect(err).NotTo(o.HaveOccurred())
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
_, stderr, err := oc.AsAdmin().WithoutNamespace().Run("apply").Args("-f", cfgFileJSON).Outputs()
o.Expect(err).To(o.HaveOccurred())
o.Expect(stderr).To(o.ContainSubstring("Invalid cluster name (.spec.clusterName): must be no more than 63 characters"))
}) | |||||
test case | openshift/openshift-tests-private | 79f2abeb-372f-4193-b3f4-bb7fd1a6585d | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-22382-ClusterDeployment.spec cannot be changed during an update[Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:fxie-High-22382-ClusterDeployment.spec cannot be changed during an update[Serial]", func() {
testCaseID := "22382"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("OCP-22382: clusterdeployment.spec does not allow edit during an update")
e2e.Logf("Make sure a provision Pod is created in the project's namespace")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "-provision-", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace()}).check(oc)
e2e.Logf("Now attempt to modify clusterdeployment.spec")
output, err := oc.AsAdmin().Run("patch").Args("cd", cdName, "--type=merge", "-p", "{\"spec\":{\"baseDomain\": \"qe1.devcluster.openshift.com\"}}").Output()
o.Expect(err).To(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("Attempted to change ClusterDeployment.Spec which is immutable"))
}) | |||||
test case | openshift/openshift-tests-private | 287891ae-8daf-42d5-bfc2-8fa763a72d70 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-22379-Medium-42721-[HiveSpec] Adopt clusters to Hive [Serial] | ['"context"', '"fmt"', '"os"', '"path/filepath"', '"strings"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/credentials"', 'cloudFormationTypes "github.com/aws/aws-sdk-go-v2/service/cloudformation/types"', '"github.com/aws/aws-sdk-go-v2/service/ec2"', '"github.com/aws/aws-sdk-go-v2/service/ec2/types"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:fxie-Medium-22379-Medium-42721-[HiveSpec] Adopt clusters to Hive [Serial]", func() {
testCaseID := "42721"
resourceNameSuffix := testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
e2e.Logf("Create ClusterImageSet")
imageSetName := "clusterimageset-" + resourceNameSuffix
clusterImageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
clusterImageSet.create(oc)
e2e.Logf("Copy AWS root credentials & pull-secret to the temporary namespace")
createAWSCreds(oc, oc.Namespace())
createPullSecret(oc, oc.Namespace())
exutil.By("Create ClusterPool, wait for it to be ready")
poolName := "clusterpool-" + resourceNameSuffix
clusterPool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "false",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 2,
maxConcurrent: 2,
hibernateAfter: "3h",
template: filepath.Join(testDataDir, "clusterpool.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
clusterPool.create(oc)
newCheck("expect", "get", asAdmin, requireNS, compare, "2", ok, ClusterInstallTimeout, []string{"ClusterPool", poolName, "-o=jsonpath={.status.ready}"}).check(oc)
e2e.Logf("Get CDs in the ClusterPool")
CDsInPool := strings.Split(strings.Trim(getCDlistfromPool(oc, poolName), "\n"), "\n")
o.Expect(len(CDsInPool)).To(o.Equal(2))
// We will use the 2 CDs as another Hive cluster and the cluster to adopt respectively
hiveCluster2, clusterToAdopt := CDsInPool[0], CDsInPool[1]
e2e.Logf("Get kubeconfig of Hive cluster 2 (%v) and the cluster to adopt (%v)", hiveCluster2, clusterToAdopt)
tmpDir2 := "/tmp/" + hiveCluster2 + "-" + getRandomString()
defer os.RemoveAll(tmpDir2)
err := os.MkdirAll(tmpDir2, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, hiveCluster2, hiveCluster2, tmpDir2)
kubeconfig2 := tmpDir2 + "/kubeconfig"
tmpDirToAdopt := "/tmp/" + clusterToAdopt + "-" + getRandomString()
defer os.RemoveAll(tmpDirToAdopt)
err = os.MkdirAll(tmpDirToAdopt, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
getClusterKubeconfig(oc, clusterToAdopt, clusterToAdopt, tmpDirToAdopt)
kubeconfigToAdopt := tmpDirToAdopt + "/kubeconfig"
e2e.Logf("Get infra ID and cluster ID of the cluster to adopt")
infraID, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("infrastructure", "cluster", "-o=jsonpath={.status.infrastructureName}", "--kubeconfig", kubeconfigToAdopt).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
clusterID, _, err := oc.AsAdmin().WithoutNamespace().Run("get").Args("clusterversion", "version", "-o=jsonpath={.spec.clusterID}", "--kubeconfig", kubeconfigToAdopt).Outputs()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Found infra ID = %v, cluster ID = %v for the cluster to adopt", infraID, clusterID)
e2e.Logf(`Set up Hive cluster 2 (%v):
1) Deploy Hive
2) Copy AWS root credentials to the default namespace
3) Copy the pull-secret to the default namespace
4) Create a Secret containing the admin kubeconfig of the cluster to adopt in the default namespace
`, hiveCluster2)
// No need to set up a new project on Hive cluster 2 as it will eventually be de-provisioned.
// We will simply use the default namespace for this cluster.
// Likewise, there is no need to clean up the resources created on Hive cluster 2.
hiveCluster2NS := "default"
origKubeconfig := oc.GetKubeconf()
origAdminKubeconfig := exutil.KubeConfigPath()
origNS := oc.Namespace()
// Defer an anonymous function so that ALL (chained) setters are executed after running the test case.
// The deferred function is executed before all defers above, which means that the oc client object
// is restored (i.e. points back to Hive cluster 1) before cleaning up resources on that cluster.
// This is what we want.
defer func(origKubeconfig, origAdminKubeconfig, origNS string) {
oc.SetKubeconf(origKubeconfig).SetAdminKubeconf(origAdminKubeconfig).SetNamespace(origNS)
}(origKubeconfig, origAdminKubeconfig, origNS)
// From this point on, the oc client object points to Hive cluster 2.
oc.SetKubeconf(kubeconfig2).SetAdminKubeconf(kubeconfig2).SetNamespace(hiveCluster2NS)
// The installHiveOperator() function deploys Hive as admin. To deploy Hive on another cluster (Hive cluster 2 here), we have 3 options:
// 1) Create a new oc client object:
// This is complicated as we cannot use the NewCLI() function, which incorporates calls to beforeEach() and afterEach()
// and those two are disallowed in g.It(). Moreover, most fields of the utils.CLI type are internal and lack setters.
// 2) Use the existing oc client object, point it to Hive cluster 2, and make sure to restore it at the end.
// This is our approach here.
// 3) Modify the existing code s.t. Hive is deployed as non-admin (as guest for ex.):
// This is again complicated as we would need to alter the existing code infrastructure to a large extent.
installHiveOperator(oc, &ns, &og, &sub, &hc, testDataDir)
createAWSCreds(oc, hiveCluster2NS)
createPullSecret(oc, hiveCluster2NS)
adminKubeconfigSecretName := "admin-kubeconfig-adopt"
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("secret", "generic", adminKubeconfigSecretName, "-n", hiveCluster2NS, "--from-file", kubeconfigToAdopt).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By(fmt.Sprintf("Adopt cluster %v into cluster %v", clusterToAdopt, hiveCluster2))
adoptCDName := clusterToAdopt + "-adopt"
adoptCD := clusterDeploymentAdopt{
name: adoptCDName,
namespace: hiveCluster2NS,
baseDomain: AWSBaseDomain,
adminKubeconfigRef: adminKubeconfigSecretName,
clusterID: clusterID,
infraID: infraID,
clusterName: adoptCDName,
manageDNS: false,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
// OCP-22379: Hive will abandon deprovision for any cluster when preserveOnDelete is true
preserveOnDelete: true,
template: filepath.Join(testDataDir, "clusterdeployment-adopt.yaml"),
}
adoptCD.create(oc)
exutil.By("Make sure the adopted CD is running on Hive cluster 2")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, 600, []string{"ClusterDeployment", adoptCDName, "-n", hiveCluster2NS, "-o=jsonpath={.status.powerState}"}).check(oc)
exutil.By("Make sure SyncSet works on Hive cluster 2")
syncSetName := "syncset-" + resourceNameSuffix
configMapName := "configmap-" + resourceNameSuffix
configMapNamespace := "namespace-" + resourceNameSuffix
syncSetResource := syncSetResource{
name: syncSetName,
namespace: hiveCluster2NS,
namespace2: configMapNamespace,
cdrefname: adoptCDName,
cmname: configMapName,
cmnamespace: configMapNamespace,
ramode: "Sync",
template: filepath.Join(testDataDir, "syncset-resource.yaml"),
}
syncSetResource.create(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, configMapName, ok, DefaultTimeout, []string{"cm", configMapName, "-n", configMapNamespace, "--kubeconfig", kubeconfigToAdopt}).check(oc)
exutil.By("Delete the adopted CD on Hive cluster 2")
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("ClusterDeployment", adoptCDName, "-n", hiveCluster2NS).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Make sure the adopted CD is gone on Hive cluster 2")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, adoptCDName, nok, DefaultTimeout, []string{"ClusterDeployment", "-n", hiveCluster2NS}).check(oc)
e2e.Logf("Make sure the cloud resources persist (here we look for the EC2 instances)")
cfg := getAWSConfig(oc, AWSRegion)
ec2Client := ec2.NewFromConfig(cfg)
describeInstancesOutput, err := ec2Client.DescribeInstances(context.Background(), &ec2.DescribeInstancesInput{
Filters: []types.Filter{
{
Name: aws.String("tag-key"),
Values: []string{"kubernetes.io/cluster/" + infraID},
},
},
MaxResults: aws.Int32(6),
})
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(len(describeInstancesOutput.Reservations)).To(o.Equal(6))
}) | |||||
test case | openshift/openshift-tests-private | cdf6447d-7153-4294-8103-4d38fcbf3458 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-High-22381-Medium-34882-High-24693-Hive additional machinepool test [Serial] | ['"encoding/json"', '"os"', '"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-High-22381-Medium-34882-High-24693-Hive additional machinepool test [Serial]", func() {
testCaseID := "34882"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("OCP-24693: Support a global pull secret override")
e2e.Logf("Granting temp user permission to create secret in Hive's namespace ...")
// This is done so that the createPullSecret function can be used on Hive's namespace
err := oc.AsAdmin().WithoutNamespace().Run("adm", "policy").Args("add-role-to-user", "edit", oc.Username(), "-n", HiveNamespace).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Creating global pull-secret ...")
defer oc.AsAdmin().Run("delete").Args("secret", "pull-secret", "-n", HiveNamespace).Execute()
createPullSecret(oc, HiveNamespace)
e2e.Logf("Patching Hiveconfig ...")
patch := `
spec:
globalPullSecretRef:
name: pull-secret`
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "-n=hive", "--type=json", "-p", `[{"op":"remove", "path": "/spec/globalPullSecretRef"}]`).Execute()
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "-n=hive", "--type=merge", "-p", patch).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
workermp.create(oc)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-22381: machinepool.spec.plaform does not allow edit")
e2e.Logf("Patch worker machinepool .spec.platform")
patchYaml := `
spec:
name: worker
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp3
type: m4.2xlarge`
output, err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("MachinePool", cdName+"-worker", "-n", oc.Namespace(), "--type", "merge", "-p", patchYaml).Output()
o.Expect(err).To(o.HaveOccurred())
o.Expect(output).To(o.ContainSubstring("field is immutable"))
e2e.Logf("Check machines type is still m4.xlarge")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "m4.xlarge", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.spec.platform.aws.type}"}).check(oc)
exutil.By("OCP-34882: Hive should be able to create additional machinepool after deleting all MachinePools")
e2e.Logf("Delete all machinepools")
cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"})
e2e.Logf("Check there are no machinepools existing")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "No resources found", ok, DefaultTimeout, []string{"MachinePool", "-n", oc.Namespace()}).check(oc)
e2e.Logf("Check there are no machinesets in remote cluster")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err = os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "No resources found", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api"}).check(oc)
e2e.Logf("Create one more infra machinepool, check it can be created")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc, objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"})
inframp.create(oc)
e2e.Logf("Check infra machinepool .status.replicas = 1 ")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "1", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
machinesetsname := getResource(oc, asAdmin, withoutNamespace, "MachinePool", cdName+"-infra", "-n", oc.Namespace(), "-o=jsonpath={.status.machineSets[?(@.replicas==1)].name}")
o.Expect(machinesetsname).NotTo(o.BeEmpty())
e2e.Logf("Remote cluster machineset list: %s", machinesetsname)
e2e.Logf("Check machineset %s can be created on remote cluster", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, machinesetsname, ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].metadata.name}"}).check(oc)
e2e.Logf("Check machineset %s is up", machinesetsname)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "1", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[?(@.spec.replicas==1)].status.availableReplicas}"}).check(oc)
e2e.Logf("Check machines is in Running status")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "Running", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=infra", "-o=jsonpath={.items[*].status.phase}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | 2266ffd9-4e48-49c0-b298-18d468ab3dd9 | Author:lwan-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-28867-Medium-41776-[HiveSpec] Hive Machinepool test for autoscale [Serial] | ['"encoding/json"', '"fmt"', '"io/ioutil"', '"os"', '"path/filepath"', '"time"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"k8s.io/apimachinery/pkg/util/wait"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("Author:lwan-NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-High-28867-Medium-41776-[HiveSpec] Hive Machinepool test for autoscale [Serial]", func() {
testCaseID := "28867"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Create worker and infra MachinePool ...")
workermachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-worker-aws.yaml")
inframachinepoolAWSTemp := filepath.Join(testDataDir, "machinepool-infra-aws.yaml")
workermp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: workermachinepoolAWSTemp,
}
inframp := machinepool{
namespace: oc.Namespace(),
clusterName: cdName,
template: inframachinepoolAWSTemp,
}
defer cleanupObjects(oc,
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-worker"},
objectTableRef{"MachinePool", oc.Namespace(), cdName + "-infra"},
)
workermp.create(oc)
inframp.create(oc)
exutil.By("Check if ClusterDeployment created successfully and become Provisioned")
//newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
exutil.By("OCP-28867: Hive supports an optional autoscaler settings instead of static replica count")
tmpDir := "/tmp/" + cdName + "-" + getRandomString()
err := os.MkdirAll(tmpDir, 0777)
o.Expect(err).NotTo(o.HaveOccurred())
defer os.RemoveAll(tmpDir)
getClusterKubeconfig(oc, cdName, oc.Namespace(), tmpDir)
kubeconfig := tmpDir + "/kubeconfig"
e2e.Logf("Patch static replicas to autoscaler")
autoScalingMax := "12"
autoScalingMin := "10"
removeConfig := "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig := fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is minimum value %s", autoScalingMin)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 3 3", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number is minReplicas %s when low workload", autoScalingMin)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 10 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't equal to minReplicas 10")
patchYaml := `
spec:
scaleDown:
enabled: true
delayAfterAdd: 10s
delayAfterDelete: 10s
delayAfterFailure: 10s
unneededTime: 10s`
e2e.Logf("Add busybox in remote cluster and check machines will scale up to maxReplicas %s", autoScalingMax)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "ClusterAutoscaler", "default", "--type", "merge", "-p", patchYaml}).check(oc)
workloadYaml := filepath.Join(testDataDir, "workload.yaml")
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "busybox", ok, DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "Deployment", "busybox", "-n", "default"}).check(oc)
e2e.Logf("Check replicas will scale up to maximum value %s", autoScalingMax)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 4 4", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number will scale up to maxReplicas %s", autoScalingMax)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 12 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't scale up to maxReplicas 12 after workload up")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "12", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
e2e.Logf("Delete busybox in remote cluster and check machines will scale down to minReplicas %s", autoScalingMin)
err = oc.AsAdmin().WithoutNamespace().Run("delete").Args("--kubeconfig="+kubeconfig, "-f", workloadYaml).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas will scale down to minimum value %s", autoScalingMin)
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "4 3 3", ok, 10*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=worker", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check machines number will scale down to minReplicas %s", autoScalingMin)
err = wait.Poll(1*time.Minute, (ClusterResumeTimeout/60)*time.Minute, func() (bool, error) {
runningMachinesNum := checkResourceNumber(oc, "Running", []string{"--kubeconfig=" + kubeconfig, "Machine", "-n", "openshift-machine-api", "-l", "machine.openshift.io/cluster-api-machine-role=worker"})
if runningMachinesNum == 10 {
return true, nil
}
return false, nil
})
exutil.AssertWaitPollNoErr(err, "machines in remote cluster doesn't scale down to minReplicas 10 after workload down")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "10", ok, 5*DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "-o=jsonpath={.status.replicas}"}).check(oc)
removeConfig = "[{\"op\": \"remove\", \"path\": \"/spec/autoscaling\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
replicas := "3"
staticConfig := fmt.Sprintf("{\"spec\": {\"replicas\": %s}}", replicas)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-worker", "-n", oc.Namespace(), "--type", "merge", "-p", staticConfig}).check(oc)
exutil.By("OCP-41776: Allow minReplicas autoscaling of MachinePools to be 0")
e2e.Logf("Check hive allow set minReplicas=0 without zone setting")
autoScalingMax = "3"
autoScalingMin = "0"
removeConfig = "[{\"op\": \"remove\", \"path\": \"/spec/replicas\"}]"
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "json", "-p", removeConfig}).check(oc)
autoscalConfig = fmt.Sprintf("{\"spec\": {\"autoscaling\": {\"maxReplicas\": %s, \"minReplicas\": %s}}}", autoScalingMax, autoScalingMin)
newCheck("expect", "patch", asAdmin, withoutNamespace, contain, "patched", ok, DefaultTimeout, []string{"MachinePool", cdName + "-infra", "-n", oc.Namespace(), "--type", "merge", "-p", autoscalConfig}).check(oc)
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 5*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
e2e.Logf("Check hive allow set minReplicas=0 within zone setting")
infra2MachinepoolYaml := `
apiVersion: hive.openshift.io/v1
kind: MachinePool
metadata:
name: ` + cdName + `-infra2
namespace: ` + oc.Namespace() + `
spec:
autoscaling:
maxReplicas: 3
minReplicas: 0
clusterDeploymentRef:
name: ` + cdName + `
labels:
node-role.kubernetes.io: infra2
node-role.kubernetes.io/infra2: ""
name: infra2
platform:
aws:
rootVolume:
iops: 100
size: 22
type: gp3
type: m4.xlarge
zones:
- ` + AWSRegion + `a
- ` + AWSRegion + `b
- ` + AWSRegion + `c
taints:
- effect: NoSchedule
key: node-role.kubernetes.io/infra2`
var filename = testCaseID + "-machinepool-infra2.yaml"
err = ioutil.WriteFile(filename, []byte(infra2MachinepoolYaml), 0644)
defer os.Remove(filename)
o.Expect(err).NotTo(o.HaveOccurred())
defer oc.AsAdmin().WithoutNamespace().Run("delete").Args("-f", filename, "--ignore-not-found").Execute()
err = oc.AsAdmin().WithoutNamespace().Run("create").Args("-f", filename).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check replicas is 0")
newCheck("expect", "get", asAdmin, withoutNamespace, compare, "0 0 0", ok, 4*DefaultTimeout, []string{"--kubeconfig=" + kubeconfig, "MachineSet", "-n", "openshift-machine-api", "-l", "hive.openshift.io/machine-pool=infra2", "-o=jsonpath={.items[*].status.replicas}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | fd014311-3709-40fa-85f5-3c47eaed2db6 | NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-High-23289-Medium-39813-Test hive reports install restarts in CD and Metric[Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-NonPreRelease-Longduration-ConnectedOnly-Author:lwan-High-23289-Medium-39813-Test hive reports install restarts in CD and Metric[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
testCaseID := "23289"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
imageSetName := cdName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
oc.SetupProject()
e2e.Logf("Create a invalid aws creds make install failed.")
e2e.Logf("Modify aws creds to invalid")
err := oc.Run("create").Args("secret", "generic", AWSCreds, "--from-literal=aws_access_key_id=test", "--from-literal=aws_secret_access_key=test", "-n", oc.Namespace()).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := cdName + "-install-config"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterDeployment with installAttemptsLimit=3...")
clusterTemp := filepath.Join(testDataDir, "clusterdeployment.yaml")
cluster := clusterDeployment{
fake: "false",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: imageSetName,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
installAttemptsLimit: 3,
template: clusterTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName})
cluster.create(oc)
exutil.By("OCP-23289: Check hive reports current number of install job retries in cluster deployment status...")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "3", ok, ClusterResumeTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.status.installRestarts}"}).check(oc)
o.Expect(checkResourceNumber(oc, cdName, []string{"pods", "-A"})).To(o.Equal(3))
exutil.By("OCP-39813: Check provision metric reporting number of install restarts...")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query := "hive_cluster_deployment_provision_underway_install_restarts"
checkResourcesMetricValue(oc, cdName, oc.Namespace(), "3", token, thanosQuerierURL, query)
}) | |||||
test case | openshift/openshift-tests-private | f3493449-a777-4411-910c-8806d2648ad7 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-27559-[HiveSpec] hive controllers can be disabled through a hiveconfig option [Disruptive] | ['"encoding/json"', '"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:mihuang-High-27559-[HiveSpec] hive controllers can be disabled through a hiveconfig option [Disruptive]", func() {
e2e.Logf("Add \"maintenanceMode: true\" in hiveconfig.spec")
defer oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", "json", "-p", `[{"op":"remove", "path": "/spec/maintenanceMode"}]`).Execute()
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig/hive", "--type", `merge`, `--patch={"spec": {"maintenanceMode": true}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check modifying is successful")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, DefaultTimeout, []string{"hiveconfig", "hive", "-o=jsonpath={.spec.maintenanceMode}"}).check(oc)
exutil.By("Check hive-clustersync and hive-controllers pods scale down, hive-operator and hiveadmission pods are not affected.")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", nok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", nok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
e2e.Logf("Patch hiveconfig.spec.maintenanceMode to false")
err = oc.AsAdmin().WithoutNamespace().Run("patch").Args("hiveconfig", "hive", "--type", "merge", "-p", `{"spec":{"maintenanceMode": false}}`).Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Verify the hive-controller and hive-clustersync pods scale up and appear")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-clustersync", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=clustersync",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-controllers", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=controller-manager",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hive-operator", ok, DefaultTimeout, []string{"pod", "--selector=control-plane=hive-operator",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "hiveadmission", ok, DefaultTimeout, []string{"pod", "--selector=app=hiveadmission",
"-n", sub.namespace, "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
testCaseID := "27559"
cdName := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment...")
cluster := clusterDeployment{
fake: "true",
name: cdName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: cdName + "-imageset",
installConfigSecret: cdName + "-install-config",
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 3,
}
defer cleanCD(oc, cluster.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster)
exutil.By("Check if ClusterDeployment created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "true", ok, ClusterInstallTimeout, []string{"ClusterDeployment", cdName, "-n", oc.Namespace(), "-o=jsonpath={.spec.installed}"}).check(oc)
}) | |||||
test case | openshift/openshift-tests-private | ff067b56-6187-43d8-ac0b-cadf71f82944 | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-44477-Medium-44474-Medium-44476-[HiveSpec] Change fields of a steady pool, all unclaimed clusters will be recreated[Serial] | ['"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"', '"github.com/aws/aws-sdk-go-v2/credentials"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:lwan-Medium-44477-Medium-44474-Medium-44476-[HiveSpec] Change fields of a steady pool, all unclaimed clusters will be recreated[Serial]", func() {
// Expose Hive metrics, and neutralize the effect after finishing the test case
needRecover, prevConfig := false, ""
defer recoverClusterMonitoring(oc, &needRecover, &prevConfig)
exposeMetrics(oc, testDataDir, &needRecover, &prevConfig)
testCaseID := "44477"
poolName := "pool-" + testCaseID
imageSetName := poolName + "-imageset"
imageSetTemp := filepath.Join(testDataDir, "clusterimageset.yaml")
imageSet := clusterImageSet{
name: imageSetName,
releaseImage: testOCPImage,
template: imageSetTemp,
}
imageSetName2 := poolName + "-imageset-2"
imageSet2 := clusterImageSet{
name: imageSetName2,
releaseImage: testOCPImage,
template: imageSetTemp,
}
exutil.By("Create ClusterImageSet...")
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName})
imageSet.create(oc)
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", imageSetName2})
imageSet2.create(oc)
exutil.By("Check if ClusterImageSet was created successfully")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName, ok, DefaultTimeout, []string{"ClusterImageSet", "-A", "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
newCheck("expect", "get", asAdmin, withoutNamespace, contain, imageSetName2, ok, DefaultTimeout, []string{"ClusterImageSet", "-A", "-o=jsonpath={.items[*].metadata.name}"}).check(oc)
oc.SetupProject()
//secrets can be accessed by pod in the same namespace, so copy pull-secret and gcp-credentials to target namespace for the clusterdeployment
exutil.By("Copy AWS platform credentials...")
createAWSCreds(oc, oc.Namespace())
exutil.By("Copy pull-secret...")
createPullSecret(oc, oc.Namespace())
exutil.By("Create Install-Config template Secret...")
installConfigTemp := filepath.Join(testDataDir, "aws-install-config.yaml")
installConfigSecretName := poolName + "-install-config-template"
installConfigSecret := installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: poolName,
region: AWSRegion,
template: installConfigTemp,
}
defer cleanupObjects(oc, objectTableRef{"secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Create ClusterPool...")
poolTemp := filepath.Join(testDataDir, "clusterpool.yaml")
pool := clusterPool{
name: poolName,
namespace: oc.Namespace(),
fake: "true",
baseDomain: AWSBaseDomain,
imageSetRef: imageSetName,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
pullSecretRef: PullSecret,
size: 2,
maxSize: 2,
runningCount: 0,
maxConcurrent: 1,
hibernateAfter: "10m",
template: poolTemp,
}
defer cleanupObjects(oc, objectTableRef{"ClusterPool", oc.Namespace(), poolName})
pool.create(oc)
exutil.By("Check if ClusterPool created successfully and become ready")
//runningCount is 0 so pool status should be standby: 2, ready: 0
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
jsonPath := "-o=jsonpath={\"reason:\"}{.status.conditions[?(@.type==\"AllClustersCurrent\")].reason}{\",status:\"}{.status.conditions[?(@.type==\"AllClustersCurrent\")].status}"
expectedResult := "reason:ClusterDeploymentsCurrent,status:True"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
field := []string{"imageSetRef", "userTags", "InstallConfigSecretTemplateRef"}
var (
caseID string
patchYaml string
jsonPathTemp string
expectedResultTemp string
)
for _, v := range field {
switch v {
case "imageSetRef":
caseID = "OCP-44476"
patchYaml = `{"spec":{"imageSetRef":{"name":"` + imageSetName2 + `"}}}`
jsonPathTemp = `-o=jsonpath={.items[?(@.spec.clusterPoolRef.poolName=="` + poolName + `")].spec.provisioning.imageSetRef.name}`
expectedResultTemp = imageSetName2 + " " + imageSetName2
case "userTags":
caseID = "OCP-44474"
patchYaml = `{"spec":{"platform":{"aws":{"userTags":{"cluster_desc":"` + poolName + `"}}}}}`
//jsonPathTemp = `-o=jsonpath={.items[?(@.spec.clusterPoolRef.poolName=="` + poolName + `")].spec.platform.aws.userTags.cluster_desc}`
//expectedResultTemp = poolName + " " + poolName
case "InstallConfigSecretTemplateRef":
caseID = "OCP-44477"
patchYaml = `{"spec":{"installConfigSecretTemplateRef":{"name":"` + installConfigSecretName + `"}}}`
default:
g.Fail("Given field" + v + " are not supported")
}
exutil.By(caseID + ": Change " + v + " field of a steady pool, all unclaimed clusters will be recreated")
e2e.Logf("oc patch ClusterPool field %s", v)
err := oc.AsAdmin().WithoutNamespace().Run("patch").Args("ClusterPool", poolName, "-n", oc.Namespace(), "-p", patchYaml, "--type=merge").Execute()
o.Expect(err).NotTo(o.HaveOccurred())
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
expectedResult = "reason:SomeClusterDeploymentsStale,status:False"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
e2e.Logf("Check ClusterPool Condition \"AllClustersCurrent\"")
expectedResult = "reason:ClusterDeploymentsCurrent,status:True"
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResult, ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), jsonPath}).check(oc)
if v == "imageSetRef" {
newCheck("expect", "get", asAdmin, withoutNamespace, contain, expectedResultTemp, ok, DefaultTimeout, []string{"ClusterDeployment", "-A", jsonPathTemp}).check(oc)
}
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "2", ok, 2*DefaultTimeout, []string{"ClusterPool", poolName, "-n", oc.Namespace(), "-o=jsonpath={.status.standby}"}).check(oc)
}
exutil.By("Check Metrics for ClusterPool...")
token, err := exutil.GetSAToken(oc)
o.Expect(err).NotTo(o.HaveOccurred())
o.Expect(token).NotTo(o.BeEmpty())
query := "hive_clusterpool_stale_clusterdeployments_deleted"
e2e.Logf("Check metric %s Value equal to 6", query)
checkResourcesMetricValue(oc, poolName, oc.Namespace(), "6", token, thanosQuerierURL, query)
}) | |||||
test case | openshift/openshift-tests-private | be0355ed-bd42-4af5-9f71-1a93bf822d8c | NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-27770-Hive should set Condition when given ClusterImageSet or image doesn't exist[Serial] | ['"fmt"', '"path/filepath"', '"github.com/aws/aws-sdk-go-v2/aws"', '"github.com/aws/aws-sdk-go-v2/config"'] | github.com/openshift/openshift-tests-private/test/extended/cluster_operator/hive/hive_aws.go | g.It("NonHyperShiftHOST-Longduration-NonPreRelease-ConnectedOnly-Author:kcui-Medium-27770-Hive should set Condition when given ClusterImageSet or image doesn't exist[Serial]", func() {
testCaseID := "27770"
cdName1 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
cdName2 := "cluster-" + testCaseID + "-" + getRandomString()[:ClusterSuffixLen]
oc.SetupProject()
exutil.By("Config cd1 Install-Config Secret...")
installConfigSecret := installConfig{
name1: cdName1 + "-install-config",
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName1,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
exutil.By("Config ClusterDeployment1...")
clusterImageSetName1 := cdName1 + "-imageset" + "-non-exist"
cluster1 := clusterDeployment{
fake: "false",
name: cdName1,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName1,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName1,
installConfigSecret: cdName1 + "-install-config",
pullSecretRef: PullSecret,
installAttemptsLimit: 1,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
}
defer cleanCD(oc, cluster1.name+"-imageset", oc.Namespace(), installConfigSecret.name1, cluster1.name)
createCD(testDataDir, testOCPImage, oc, oc.Namespace(), installConfigSecret, cluster1)
exutil.By("Creating cd2 install-config Secret ...")
installConfigSecretName := cdName2 + "-install-config"
installConfigSecret = installConfig{
name1: installConfigSecretName,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
name2: cdName2,
region: AWSRegion,
template: filepath.Join(testDataDir, "aws-install-config.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"Secret", oc.Namespace(), installConfigSecretName})
installConfigSecret.create(oc)
exutil.By("Creating cd2 ClusterImageSet with WrongReleaseImage...")
clusterImageSetName2 := cdName2 + "-imageset"
WrongReleaseImage := "registry.ci.openshift.org/ocp/release:4.13.0-0.nightly-2023-02-26-081527-non-exist"
imageSet := clusterImageSet{
name: clusterImageSetName2,
releaseImage: WrongReleaseImage,
template: filepath.Join(testDataDir, "clusterimageset.yaml"),
}
defer cleanupObjects(oc, objectTableRef{"ClusterImageSet", "", clusterImageSetName2})
imageSet.create(oc)
exutil.By("Creating cd2 with an incomplete pull-secret ...")
cluster2 := clusterDeployment{
fake: "false",
name: cdName2,
namespace: oc.Namespace(),
baseDomain: AWSBaseDomain,
clusterName: cdName2,
platformType: "aws",
credRef: AWSCreds,
region: AWSRegion,
imageSetRef: clusterImageSetName2,
installConfigSecret: installConfigSecretName,
pullSecretRef: PullSecret,
template: filepath.Join(testDataDir, "clusterdeployment.yaml"),
installAttemptsLimit: 1,
}
defer cleanupObjects(oc, objectTableRef{"ClusterDeployment", oc.Namespace(), cdName2})
cluster2.create(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the message 'ClusterImageSet clusterImageSetName is not available'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, fmt.Sprintf("ClusterImageSet %s is not available", clusterImageSetName1), ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].message}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the reason 'ClusterImageSetNotFound'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "ClusterImageSetNotFound", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].reason}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'RequirementsMet',return the status 'False'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "False", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"RequirementsMet\")].status}'"}).check(oc)
exutil.By("Check cd1 conditions with type 'ClusterImageSetNotFound', return no output")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "", ok, DefaultTimeout, []string{"ClusterDeployment", cdName1, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"ClusterImageSetNotFound\")]}'"}).check(oc)
exutil.By("Check pod pf cd2, return the status 'failed with Init:ImagePullBackOff'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "Init:ImagePullBackOff", ok, DefaultTimeout, []string{"pod", "-n", oc.Namespace(), "--selector", "hive.openshift.io/imageset=true", "--selector", fmt.Sprintf("hive.openshift.io/cluster-deployment-name=%s", cdName2), "--no-headers"}).check(oc)
exutil.By("Check cd2 conditions with type 'installImagesNotResolved',return the reason 'JobToResolveImagesFailed'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "JobToResolveImagesFailed", ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"InstallImagesNotResolved\")].reason}'"}).check(oc)
exutil.By("Check cd2 conditions with type 'RequirementsMet',return the status 'True'")
newCheck("expect", "get", asAdmin, withoutNamespace, contain, "True", ok, DefaultTimeout, []string{"ClusterDeployment", cdName2, "-n", oc.Namespace(), "-o=jsonpath='{.status.conditions[?(@.type == \"InstallImagesNotResolved\")].status}'"}).check(oc)
}) |
Subsets and Splits