Datasets:
edg3
/
experiment_software_csharp_12_net_8

Dataset card Data Studio Files Community
1
title
stringlengths
3
46
content
stringlengths
0
1.6k
22:250
22:251
You can verify that your images have been correctly loaded by listing all the images loaded in Minikube:
22:252
minikube image ls
22:253
22:254
Now, we need to define a .yaml Kubernetes configuration file with two deployments and a service that forwards communications to grpcmicroservice, which is the only microservice acting as a server. Let’s call it minikubedeploy.yaml.
22:255
The definition of the grpcmicroservice deployment is straightforward:
22:256
apiVersion: apps/v1
22:257
kind: Deployment
22:258
metadata:
22:259
name: grpcmicroservice
22:260
labels:
22:261
app: statistics
22:262
spec:
22:263
selector:
22:264
matchLabels:
22:265
app: statistics
22:266
role: worker
22:267
replicas: 1
22:268
template:
22:269
metadata:
22:270
labels:
22:271
app: statistics
22:272
role: worker
22:273
spec:
22:274
containers:
22:275
- name: grpcmicroservice
22:276
image: grpcmicroservice:latest
22:277
imagePullPolicy: Never
22:278
resources:
22:279
requests:
22:280
cpu: 10m
22:281
memory: 10Mi
22:282
env:
22:283
- name: ASPNETCORE_HTTP_PORTS
22:284
value: `8080`
22:285
ports:
22:286
- containerPort: 8080
22:287
name: http
22:288
22:289
The code above requires just one replica, but you can experiment with two or three replicas. The ASPNETCORE_HTTP_PORTS environment variable is a standard ASP.NET setting that informs Kestrel on the HTTP port where to listen.
22:290
22:291
The imagePullPolicy: Never setting specifies the image caching policy within the Kubernetes cluster. It prevents Minikube from trying to download a fresher version of the image from the original source into its cache. We need this setting, since there is no “original source” containing our image because we uploaded the image directly to the Minikube cache with the minikube image load command.
22:292
You must always specify this setting when images are not available in a shared image repository but are uploaded directly to the Minikube cache from the Docker Desktop local repository. Shared images, instead, do not need to be uploaded manually in the Minikube cache but can be simply referenced with their full URL in the Kubernetes .yaml file.
22:293
22:294
All other settings are quite standard.
22:295
The definition of the fakesource deployment is completely analogous but doesn’t contain information about the container ports, since this microservice doesn’t act as a server:
22:296
apiVersion: apps/v1
22:297
kind: Deployment
22:298
metadata:
22:299
name: fakesource
22:300
labels:
22:301
app: sale
22:302
spec:
22:303
selector:
22:304
matchLabels:
22:305
app: sales
22:306
role: source
22:307
replicas: 1
22:308
template:
22:309
metadata:
22:310
labels:
22:311
app: sales
22:312
role: source
22:313
spec:
22:314
containers:
22:315
- name: fakesource
22:316
image: fakesource:latest
22:317
imagePullPolicy: Never
22:318
resources:
22:319
requests:
22:320
cpu: 10m
22:321
memory: 10Mi
22:322
22:323
The definition of the service that forwards communications to grpcmicroservice is quite standard:
22:324
apiVersion: v1
22:325
kind: Service
22:326
metadata:
22:327
name: grpcmicroservice
22:328
labels:
22:329
app: contract
22:330
role: worker
22:331
spec:
22:332
ports:
22:333
- port: 8080
22:334
name: http
22:335
protocol: TCP
22:336
targetPort: 8080
22:337
selector:
22:338
app: statistics
22:339
role: worker
22:340
22:341
You must pay attention only to the port numbers that must be coherent in all settings and to the service name, since they will be used in the URLs of all communication to the grpcmicroservice.
22:342
22:343
If the service names match the hostnames in the Docker virtual network, the URLs will work both in Kubernetes and the Docker virtual network. So, you don’t need to modify any code or configuration to adapt the code that runs in the Docker virtual network to Minikube or any other Kubernetes clusters.
22:344
22:345
The whole minkubedeploy.yaml file is available in the ch22 folder of the GitHub repository associated with the book.
22:346
Now, let’s open a Windows prompt in the folder that contains the minkubedeploy.yaml file, issuing the command below that will load the application configuration in the Minikube cluster:
22:347
kubectl create -f minkubedeploy.yaml
22:348
22:349
Then, issue the kubectl get deployment command to verify that all deployments have been correctly defined and are running.