Upload 3 files
Browse files- Dockerfile +23 -22
- README.md +11 -11
- sync_data.sh +127 -123
Dockerfile
CHANGED
@@ -1,22 +1,23 @@
|
|
1 |
-
FROM chaoszhu/easynode AS base
|
2 |
-
|
3 |
-
RUN apk add --no-cache python3 python3-dev py3-pip git
|
4 |
-
|
5 |
-
RUN mkdir -p /easynode/app/logs && \
|
6 |
-
mkdir -p /easynode/app/db && \
|
7 |
-
chown -R node:node /easynode/app && \
|
8 |
-
chmod -R 755 /easynode/app
|
9 |
-
|
10 |
-
ENV VIRTUAL_ENV=/easynode/venv
|
11 |
-
RUN python3 -m venv $VIRTUAL_ENV
|
12 |
-
ENV PATH="$VIRTUAL_ENV/bin:$PATH"
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
|
|
|
1 |
+
FROM chaoszhu/easynode AS base
|
2 |
+
|
3 |
+
RUN apk add --no-cache python3 python3-dev py3-pip git
|
4 |
+
|
5 |
+
RUN mkdir -p /easynode/app/logs && \
|
6 |
+
mkdir -p /easynode/app/db && \
|
7 |
+
chown -R node:node /easynode/app && \
|
8 |
+
chmod -R 755 /easynode/app
|
9 |
+
|
10 |
+
ENV VIRTUAL_ENV=/easynode/venv
|
11 |
+
RUN python3 -m venv $VIRTUAL_ENV
|
12 |
+
ENV PATH="$VIRTUAL_ENV/bin:$PATH"
|
13 |
+
ENV PORT=8083
|
14 |
+
|
15 |
+
RUN pip install --no-cache-dir huggingface_hub
|
16 |
+
|
17 |
+
USER node
|
18 |
+
WORKDIR /easynode/app
|
19 |
+
|
20 |
+
COPY --chown=node:node sync_data.sh /easynode/app/
|
21 |
+
RUN chmod +x /easynode/app/sync_data.sh
|
22 |
+
|
23 |
+
CMD ["/bin/sh", "-c", "./sync_data.sh & npm run start"]
|
README.md
CHANGED
@@ -1,11 +1,11 @@
|
|
1 |
-
---
|
2 |
-
title: Wanlian
|
3 |
-
emoji: 📉
|
4 |
-
colorFrom: yellow
|
5 |
-
colorTo: gray
|
6 |
-
sdk: docker
|
7 |
-
pinned: false
|
8 |
-
app_port:
|
9 |
-
---
|
10 |
-
|
11 |
-
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
1 |
+
---
|
2 |
+
title: Wanlian
|
3 |
+
emoji: 📉
|
4 |
+
colorFrom: yellow
|
5 |
+
colorTo: gray
|
6 |
+
sdk: docker
|
7 |
+
pinned: false
|
8 |
+
app_port: 8083
|
9 |
+
---
|
10 |
+
|
11 |
+
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
sync_data.sh
CHANGED
@@ -1,124 +1,128 @@
|
|
1 |
-
#!/bin/sh
|
2 |
-
|
3 |
-
# 检查环境变量
|
4 |
-
if [ -z "$HF_TOKEN" ] || [ -z "$DATASET_ID" ]; then
|
5 |
-
echo "Starting without backup functionality - missing HF_TOKEN or DATASET_ID"
|
6 |
-
exit 1
|
7 |
-
fi
|
8 |
-
|
9 |
-
# 激活虚拟环境
|
10 |
-
. /easynode/venv/bin/activate
|
11 |
-
|
12 |
-
# 上传备份
|
13 |
-
cat > /easynode/app/hf_sync.py << 'EOL'
|
14 |
-
from huggingface_hub import HfApi
|
15 |
-
import sys
|
16 |
-
import os
|
17 |
-
import tarfile
|
18 |
-
import tempfile
|
19 |
-
|
20 |
-
def manage_backups(api, repo_id, max_files=50):
|
21 |
-
files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
|
22 |
-
backup_files = [f for f in files if f.startswith('easynode_backup_') and f.endswith('.tar.gz')]
|
23 |
-
backup_files.sort()
|
24 |
-
|
25 |
-
if len(backup_files) >= max_files:
|
26 |
-
files_to_delete = backup_files[:(len(backup_files) - max_files + 1)]
|
27 |
-
for file_to_delete in files_to_delete:
|
28 |
-
try:
|
29 |
-
api.delete_file(path_in_repo=file_to_delete, repo_id=repo_id, repo_type="dataset")
|
30 |
-
print(f'Deleted old backup: {file_to_delete}')
|
31 |
-
except Exception as e:
|
32 |
-
print(f'Error deleting {file_to_delete}: {str(e)}')
|
33 |
-
|
34 |
-
def upload_backup(file_path, file_name, token, repo_id):
|
35 |
-
api = HfApi(token=token)
|
36 |
-
try:
|
37 |
-
api.upload_file(
|
38 |
-
path_or_fileobj=file_path,
|
39 |
-
path_in_repo=file_name,
|
40 |
-
repo_id=repo_id,
|
41 |
-
repo_type="dataset"
|
42 |
-
)
|
43 |
-
print(f"Successfully uploaded {file_name}")
|
44 |
-
|
45 |
-
manage_backups(api, repo_id)
|
46 |
-
except Exception as e:
|
47 |
-
print(f"Error uploading file: {str(e)}")
|
48 |
-
|
49 |
-
# 下载最新备份
|
50 |
-
def download_latest_backup(token, repo_id):
|
51 |
-
try:
|
52 |
-
api = HfApi(token=token)
|
53 |
-
files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
|
54 |
-
backup_files = [f for f in files if f.startswith('easynode_backup_') and f.endswith('.tar.gz')]
|
55 |
-
|
56 |
-
if not backup_files:
|
57 |
-
print("No backup files found")
|
58 |
-
return
|
59 |
-
|
60 |
-
latest_backup = sorted(backup_files)[-1]
|
61 |
-
|
62 |
-
with tempfile.TemporaryDirectory() as temp_dir:
|
63 |
-
filepath = api.hf_hub_download(
|
64 |
-
repo_id=repo_id,
|
65 |
-
filename=latest_backup,
|
66 |
-
repo_type="dataset",
|
67 |
-
local_dir=temp_dir
|
68 |
-
)
|
69 |
-
|
70 |
-
if filepath and os.path.exists(filepath):
|
71 |
-
with tarfile.open(filepath, 'r:gz') as tar:
|
72 |
-
tar.extractall('/easynode/app/')
|
73 |
-
print(f"Successfully restored backup from {latest_backup}")
|
74 |
-
|
75 |
-
except Exception as e:
|
76 |
-
print(f"Error downloading backup: {str(e)}")
|
77 |
-
|
78 |
-
if __name__ == "__main__":
|
79 |
-
action = sys.argv[1]
|
80 |
-
token = sys.argv[2]
|
81 |
-
repo_id = sys.argv[3]
|
82 |
-
|
83 |
-
if action == "upload":
|
84 |
-
file_path = sys.argv[4]
|
85 |
-
file_name = sys.argv[5]
|
86 |
-
upload_backup(file_path, file_name, token, repo_id)
|
87 |
-
elif action == "download":
|
88 |
-
download_latest_backup(token, repo_id)
|
89 |
-
EOL
|
90 |
-
|
91 |
-
# 首次启动时下载最新备份
|
92 |
-
echo "Downloading latest backup from HuggingFace..."
|
93 |
-
python hf_sync.py download "${HF_TOKEN}" "${DATASET_ID}"
|
94 |
-
|
95 |
-
# 同步函数
|
96 |
-
sync_data() {
|
97 |
-
while true; do
|
98 |
-
echo "Starting sync process at $(date)"
|
99 |
-
|
100 |
-
if [ -d "/easynode/app/db" ]; then
|
101 |
-
timestamp=$(date +%Y%m%d_%H%M%S)
|
102 |
-
backup_file="easynode_backup_${timestamp}.tar.gz"
|
103 |
-
|
104 |
-
tar -czf "/tmp/${backup_file}" db/
|
105 |
-
|
106 |
-
echo "Uploading backup to HuggingFace..."
|
107 |
-
python hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
|
108 |
-
|
109 |
-
rm -f "/tmp/${backup_file}"
|
110 |
-
else
|
111 |
-
echo "Data directory does not exist yet, waiting for next sync..."
|
112 |
-
fi
|
113 |
-
|
114 |
-
SYNC_INTERVAL=${SYNC_INTERVAL:-7200}
|
115 |
-
echo "Next sync in ${SYNC_INTERVAL} seconds..."
|
116 |
-
sleep $SYNC_INTERVAL
|
117 |
-
done
|
118 |
-
}
|
119 |
-
|
120 |
-
# 启动同步进程
|
121 |
-
sync_data &
|
122 |
-
|
123 |
-
#
|
|
|
|
|
|
|
|
|
124 |
exec npm run start
|
|
|
1 |
+
#!/bin/sh
|
2 |
+
|
3 |
+
# 检查环境变量
|
4 |
+
if [ -z "$HF_TOKEN" ] || [ -z "$DATASET_ID" ]; then
|
5 |
+
echo "Starting without backup functionality - missing HF_TOKEN or DATASET_ID"
|
6 |
+
exit 1
|
7 |
+
fi
|
8 |
+
|
9 |
+
# 激活虚拟环境
|
10 |
+
. /easynode/venv/bin/activate
|
11 |
+
|
12 |
+
# 上传备份
|
13 |
+
cat > /easynode/app/hf_sync.py << 'EOL'
|
14 |
+
from huggingface_hub import HfApi
|
15 |
+
import sys
|
16 |
+
import os
|
17 |
+
import tarfile
|
18 |
+
import tempfile
|
19 |
+
|
20 |
+
def manage_backups(api, repo_id, max_files=50):
|
21 |
+
files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
|
22 |
+
backup_files = [f for f in files if f.startswith('easynode_backup_') and f.endswith('.tar.gz')]
|
23 |
+
backup_files.sort()
|
24 |
+
|
25 |
+
if len(backup_files) >= max_files:
|
26 |
+
files_to_delete = backup_files[:(len(backup_files) - max_files + 1)]
|
27 |
+
for file_to_delete in files_to_delete:
|
28 |
+
try:
|
29 |
+
api.delete_file(path_in_repo=file_to_delete, repo_id=repo_id, repo_type="dataset")
|
30 |
+
print(f'Deleted old backup: {file_to_delete}')
|
31 |
+
except Exception as e:
|
32 |
+
print(f'Error deleting {file_to_delete}: {str(e)}')
|
33 |
+
|
34 |
+
def upload_backup(file_path, file_name, token, repo_id):
|
35 |
+
api = HfApi(token=token)
|
36 |
+
try:
|
37 |
+
api.upload_file(
|
38 |
+
path_or_fileobj=file_path,
|
39 |
+
path_in_repo=file_name,
|
40 |
+
repo_id=repo_id,
|
41 |
+
repo_type="dataset"
|
42 |
+
)
|
43 |
+
print(f"Successfully uploaded {file_name}")
|
44 |
+
|
45 |
+
manage_backups(api, repo_id)
|
46 |
+
except Exception as e:
|
47 |
+
print(f"Error uploading file: {str(e)}")
|
48 |
+
|
49 |
+
# 下载最新备份
|
50 |
+
def download_latest_backup(token, repo_id):
|
51 |
+
try:
|
52 |
+
api = HfApi(token=token)
|
53 |
+
files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
|
54 |
+
backup_files = [f for f in files if f.startswith('easynode_backup_') and f.endswith('.tar.gz')]
|
55 |
+
|
56 |
+
if not backup_files:
|
57 |
+
print("No backup files found")
|
58 |
+
return
|
59 |
+
|
60 |
+
latest_backup = sorted(backup_files)[-1]
|
61 |
+
|
62 |
+
with tempfile.TemporaryDirectory() as temp_dir:
|
63 |
+
filepath = api.hf_hub_download(
|
64 |
+
repo_id=repo_id,
|
65 |
+
filename=latest_backup,
|
66 |
+
repo_type="dataset",
|
67 |
+
local_dir=temp_dir
|
68 |
+
)
|
69 |
+
|
70 |
+
if filepath and os.path.exists(filepath):
|
71 |
+
with tarfile.open(filepath, 'r:gz') as tar:
|
72 |
+
tar.extractall('/easynode/app/')
|
73 |
+
print(f"Successfully restored backup from {latest_backup}")
|
74 |
+
|
75 |
+
except Exception as e:
|
76 |
+
print(f"Error downloading backup: {str(e)}")
|
77 |
+
|
78 |
+
if __name__ == "__main__":
|
79 |
+
action = sys.argv[1]
|
80 |
+
token = sys.argv[2]
|
81 |
+
repo_id = sys.argv[3]
|
82 |
+
|
83 |
+
if action == "upload":
|
84 |
+
file_path = sys.argv[4]
|
85 |
+
file_name = sys.argv[5]
|
86 |
+
upload_backup(file_path, file_name, token, repo_id)
|
87 |
+
elif action == "download":
|
88 |
+
download_latest_backup(token, repo_id)
|
89 |
+
EOL
|
90 |
+
|
91 |
+
# 首次启动时下载最新备份
|
92 |
+
echo "Downloading latest backup from HuggingFace..."
|
93 |
+
python hf_sync.py download "${HF_TOKEN}" "${DATASET_ID}"
|
94 |
+
|
95 |
+
# 同步函数
|
96 |
+
sync_data() {
|
97 |
+
while true; do
|
98 |
+
echo "Starting sync process at $(date)"
|
99 |
+
|
100 |
+
if [ -d "/easynode/app/db" ]; then
|
101 |
+
timestamp=$(date +%Y%m%d_%H%M%S)
|
102 |
+
backup_file="easynode_backup_${timestamp}.tar.gz"
|
103 |
+
|
104 |
+
tar -czf "/tmp/${backup_file}" db/
|
105 |
+
|
106 |
+
echo "Uploading backup to HuggingFace..."
|
107 |
+
python hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
|
108 |
+
|
109 |
+
rm -f "/tmp/${backup_file}"
|
110 |
+
else
|
111 |
+
echo "Data directory does not exist yet, waiting for next sync..."
|
112 |
+
fi
|
113 |
+
|
114 |
+
SYNC_INTERVAL=${SYNC_INTERVAL:-7200}
|
115 |
+
echo "Next sync in ${SYNC_INTERVAL} seconds..."
|
116 |
+
sleep $SYNC_INTERVAL
|
117 |
+
done
|
118 |
+
}
|
119 |
+
|
120 |
+
# 启动同步进程
|
121 |
+
sync_data &
|
122 |
+
|
123 |
+
# 确保使用正确的端口
|
124 |
+
export PORT=${PORT:-8083}
|
125 |
+
echo "Starting application on port: $PORT"
|
126 |
+
|
127 |
+
# 启动主应用
|
128 |
exec npm run start
|