Update sync_data.sh
Browse files- sync_data.sh +226 -246
sync_data.sh
CHANGED
|
@@ -1,356 +1,336 @@
|
|
| 1 |
#!/bin/bash
|
| 2 |
|
| 3 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
if [[ -z "$HF_TOKEN" ]] || [[ -z "$DATASET_ID" ]]; then
|
| 5 |
-
echo "Starting
|
| 6 |
-
#
|
| 7 |
-
|
| 8 |
-
#
|
| 9 |
-
# 注意:如果不需要 Aria2,应移除下面的 Aria2 启动或确保 Dockerfile 的 CMD 也移除了 Aria2
|
| 10 |
-
# aria2c [YOUR_ARIA2_OPTIONS_HERE] & # 根据需要添加 Aria2 参数
|
| 11 |
exec /opt/cloudreve/cloudreve -c /opt/cloudreve/config.ini
|
| 12 |
-
exit 0
|
| 13 |
fi
|
| 14 |
|
| 15 |
-
#
|
| 16 |
echo "Activating Python venv..."
|
| 17 |
source /opt/venv/bin/activate
|
| 18 |
|
| 19 |
-
#
|
| 20 |
CLOUDREVE_DIR="/opt/cloudreve"
|
| 21 |
BACKUP_PREFIX="cloudreve_backup"
|
| 22 |
CONFIG_FILE_PATH="/opt/cloudreve/config.ini"
|
| 23 |
DB_FILE_PATH="/opt/cloudreve/cloudreve.db"
|
| 24 |
EXECUTABLE_PATH="/opt/cloudreve/cloudreve"
|
| 25 |
|
| 26 |
-
# --- Python 函数定义 ---
|
| 27 |
|
| 28 |
-
# Python
|
|
|
|
|
|
|
| 29 |
upload_backup() {
|
| 30 |
-
file_path="$1"
|
| 31 |
-
file_name="$2"
|
| 32 |
-
token="$HF_TOKEN"
|
| 33 |
-
repo_id="$DATASET_ID"
|
| 34 |
|
| 35 |
echo "Preparing to upload backup file: $file_path as $file_name to Dataset: $repo_id"
|
| 36 |
|
|
|
|
|
|
|
| 37 |
python3 -c "
|
| 38 |
from huggingface_hub import HfApi
|
| 39 |
-
import sys
|
| 40 |
-
import os
|
| 41 |
print(f'HF_TOKEN is set: {os.environ.get(\"HF_TOKEN\") is not None}')
|
| 42 |
print(f'DATASET_ID is set: {os.environ.get(\"DATASET_ID\") is not None}')
|
| 43 |
-
def manage_backups(api, repo_id_val, max_files=
|
| 44 |
print('Managing old backups...')
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
|
|
|
|
|
|
|
|
|
| 61 |
api = HfApi(token='$token')
|
| 62 |
try:
|
| 63 |
-
repo_id_val = os.environ.get('DATASET_ID')
|
| 64 |
-
if not repo_id_val:
|
| 65 |
-
raise ValueError('DATASET_ID environment variable is not set.')
|
| 66 |
print(f'Uploading file: $file_path to {repo_id_val} as $file_name')
|
| 67 |
-
api.upload_file(
|
| 68 |
-
path_or_fileobj='$file_path',
|
| 69 |
-
path_in_repo='$file_name',
|
| 70 |
-
repo_id=repo_id_val,
|
| 71 |
-
repo_type='dataset'
|
| 72 |
-
)
|
| 73 |
print(f'Successfully uploaded $file_name')
|
| 74 |
manage_backups(api, repo_id_val)
|
| 75 |
except Exception as e:
|
| 76 |
print(f'Error uploading file: {str(e)}')
|
| 77 |
-
sys.exit(1) #
|
| 78 |
"
|
|
|
|
|
|
|
| 79 |
}
|
| 80 |
|
| 81 |
-
# Python
|
| 82 |
download_latest_backup() {
|
| 83 |
-
|
| 84 |
-
|
|
|
|
| 85 |
|
| 86 |
-
|
|
|
|
| 87 |
|
| 88 |
-
|
| 89 |
-
from huggingface_hub import HfApi, hf_hub_download
|
| 90 |
-
import sys
|
| 91 |
-
import os
|
| 92 |
-
import tarfile
|
| 93 |
-
import tempfile
|
| 94 |
-
import shutil
|
| 95 |
-
import subprocess
|
| 96 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
print(f'HF_TOKEN is set: {os.environ.get(\"HF_TOKEN\") is not None}')
|
| 98 |
print(f'DATASET_ID is set: {os.environ.get(\"DATASET_ID\") is not None}')
|
|
|
|
| 99 |
|
| 100 |
api = HfApi(token='$token')
|
| 101 |
try:
|
| 102 |
-
repo_id_val = os.environ.get('DATASET_ID')
|
| 103 |
-
if not repo_id_val:
|
| 104 |
-
raise ValueError('DATASET_ID environment variable is not set.')
|
| 105 |
|
| 106 |
print(f'Listing files in Dataset: {repo_id_val}')
|
| 107 |
files = api.list_repo_files(repo_id=repo_id_val, repo_type='dataset')
|
| 108 |
-
backup_files = [f for f in files if f.startswith('$BACKUP_PREFIX') and f.endswith('.tar.gz')]
|
| 109 |
|
| 110 |
if not backup_files:
|
| 111 |
print('No backup files found in the Dataset. Skipping restore.')
|
| 112 |
-
#
|
| 113 |
-
sys.exit(0)
|
| 114 |
|
| 115 |
-
latest_backup =
|
| 116 |
print(f'Latest backup file found: {latest_backup}')
|
| 117 |
|
| 118 |
with tempfile.TemporaryDirectory() as temp_dir:
|
| 119 |
print(f'Downloading {latest_backup} to temporary directory {temp_dir}...')
|
| 120 |
try:
|
| 121 |
-
|
| 122 |
-
repo_id=repo_id_val,
|
| 123 |
-
filename=latest_backup,
|
| 124 |
-
repo_type='dataset',
|
| 125 |
-
local_dir=temp_dir,
|
| 126 |
-
token=os.environ.get('HF_TOKEN') # Pass token explicitly if needed
|
| 127 |
-
)
|
| 128 |
except Exception as download_error:
|
| 129 |
-
|
| 130 |
-
# Attempt to list files again for debugging
|
| 131 |
-
try:
|
| 132 |
-
print('Attempting to list repo files again for debugging...')
|
| 133 |
-
files_debug = api.list_repo_files(repo_id=repo_id_val, repo_type='dataset')
|
| 134 |
-
print(f'Files found (debug): {files_debug}')
|
| 135 |
-
except Exception as list_error:
|
| 136 |
-
print(f'Error listing files during debug: {list_error}')
|
| 137 |
-
sys.exit(1) # Exit if download fails
|
| 138 |
-
|
| 139 |
|
| 140 |
if filepath and os.path.exists(filepath):
|
| 141 |
print(f'Successfully downloaded backup to temporary directory: {filepath}')
|
| 142 |
-
|
| 143 |
-
# Files/Dirs to restore (relative paths within CLOUDREVE_DIR)
|
| 144 |
items_to_restore = ['cloudreve', 'cloudreve.db', 'config.ini']
|
| 145 |
-
|
| 146 |
-
# Ensure target directory exists
|
| 147 |
os.makedirs(\"$CLOUDREVE_DIR\", exist_ok=True)
|
|
|
|
| 148 |
|
| 149 |
-
|
| 150 |
-
subprocess.run(['ls', '-lA', \"$CLOUDREVE_DIR\"], check=False) # Use -A to show hidden files
|
| 151 |
-
|
| 152 |
-
# --- Safer Restore Logic ---
|
| 153 |
-
# 1. Extract backup to a temporary location first
|
| 154 |
-
extract_temp_dir = os.path.join(temp_dir, 'extracted_backup')
|
| 155 |
-
os.makedirs(extract_temp_dir, exist_ok=True)
|
| 156 |
print(f'Extracting backup archive: {filepath} to {extract_temp_dir}')
|
| 157 |
try:
|
| 158 |
-
|
| 159 |
-
with tarfile.open(filepath, 'r:gz') as tar:
|
| 160 |
-
tar.extractall(extract_temp_dir)
|
| 161 |
print('Extraction complete.')
|
| 162 |
-
except
|
| 163 |
-
print(f'Error reading tar file: {tar_err}')
|
| 164 |
-
sys.exit(1)
|
| 165 |
-
except Exception as extract_err:
|
| 166 |
-
print(f'Error during extraction: {extract_err}')
|
| 167 |
-
sys.exit(1)
|
| 168 |
-
|
| 169 |
|
| 170 |
-
# 2. Check if essential files exist in the extracted backup
|
| 171 |
essential_files_present = True
|
| 172 |
-
print(f'Checking for essential items in extracted backup at {extract_temp_dir}:')
|
| 173 |
for item in items_to_restore:
|
| 174 |
-
|
| 175 |
-
|
| 176 |
-
|
| 177 |
-
|
| 178 |
-
essential_files_present = False
|
| 179 |
-
# break # Stop checking (optional, maybe list all missing?)
|
| 180 |
-
|
| 181 |
if not essential_files_present:
|
| 182 |
-
print('Listing contents of extraction directory for debugging:')
|
| 183 |
-
subprocess.run(['ls', '-lA', extract_temp_dir], check=False)
|
| 184 |
sys.exit(1) # Abort if essential files are missing
|
| 185 |
|
| 186 |
-
# 3. Delete existing items in the target directory
|
| 187 |
print(f'Deleting existing items in $CLOUDREVE_DIR before restoring...')
|
| 188 |
for item in items_to_restore:
|
| 189 |
target_path = os.path.join(\"$CLOUDREVE_DIR\", item)
|
| 190 |
if os.path.exists(target_path):
|
| 191 |
try:
|
| 192 |
-
if os.path.isdir(target_path) and not os.path.islink(target_path):
|
| 193 |
-
|
| 194 |
-
|
| 195 |
-
else:
|
| 196 |
-
print(f'Deleting file/link: {target_path}')
|
| 197 |
-
os.remove(target_path)
|
| 198 |
-
except OSError as e:
|
| 199 |
-
print(f'Error deleting {target_path}: {e}. Continuing...')
|
| 200 |
-
|
| 201 |
|
| 202 |
-
# 4. Move extracted items to the target directory
|
| 203 |
print(f'Moving extracted items from {extract_temp_dir} to $CLOUDREVE_DIR...')
|
| 204 |
for item in items_to_restore:
|
| 205 |
source_path = os.path.join(extract_temp_dir, item)
|
| 206 |
-
# Check if source exists before moving (should exist due to check above, but belt-and-suspenders)
|
| 207 |
if os.path.exists(source_path):
|
| 208 |
target_path = os.path.join(\"$CLOUDREVE_DIR\", item)
|
| 209 |
-
try:
|
| 210 |
-
|
| 211 |
-
|
| 212 |
-
except Exception as move_err:
|
| 213 |
-
print(f'Error moving {item}: {move_err}')
|
| 214 |
-
# Decide if this is critical, maybe exit? For now, print and continue.
|
| 215 |
-
else:
|
| 216 |
-
# This case should ideally not happen if the check passed
|
| 217 |
-
print(f'Warning: Source item {item} disappeared before move.')
|
| 218 |
-
|
| 219 |
|
| 220 |
print(f'Successfully restored backup from {latest_backup}')
|
| 221 |
-
print('Listing contents after restore:')
|
| 222 |
-
|
|
|
|
|
|
|
|
|
|
| 223 |
else:
|
| 224 |
print(f'Error: Downloaded file path "{filepath}" does not exist or download failed.')
|
| 225 |
-
sys.exit(1)
|
| 226 |
|
| 227 |
-
except ValueError as ve:
|
| 228 |
-
|
| 229 |
-
sys.exit(1)
|
| 230 |
-
except Exception as e:
|
| 231 |
-
print(f'Error during backup download/restore: {str(e)}')
|
| 232 |
-
# Print traceback for more details
|
| 233 |
-
import traceback
|
| 234 |
-
traceback.print_exc()
|
| 235 |
-
sys.exit(1) # Exit on error
|
| 236 |
"
|
|
|
|
|
|
|
| 237 |
}
|
| 238 |
|
| 239 |
-
|
| 240 |
-
|
| 241 |
-
|
| 242 |
-
# **新增:等待 Cloudreve 启动完成**
|
| 243 |
-
echo "Background Sync Process: Initializing..."
|
| 244 |
-
echo "Waiting for Cloudreve service to become available on port 5212..."
|
| 245 |
-
# 循环检查端口 5212 是否在监听,nc -z 返回 0 表示成功
|
| 246 |
-
while ! nc -z 127.0.0.1 5212; do
|
| 247 |
-
echo "Cloudreve not ready yet (port 5212 not open), sleeping for 5 seconds..."
|
| 248 |
-
sleep 5
|
| 249 |
-
done
|
| 250 |
-
echo "Cloudreve service detected on port 5212. Starting main sync loop."
|
| 251 |
-
# **等待逻辑结束**
|
| 252 |
-
|
| 253 |
-
# --- 原有的 sync_data 循环逻辑开始 ---
|
| 254 |
-
# echo "Background Sync Process Started" # 可以移除此行,上面的初始化消息更清晰
|
| 255 |
while true; do
|
| 256 |
-
|
| 257 |
-
#
|
| 258 |
-
|
| 259 |
-
|
| 260 |
-
|
| 261 |
-
|
| 262 |
-
|
| 263 |
-
|
| 264 |
-
|
| 265 |
-
|
| 266 |
-
# Define backup path and name
|
| 267 |
-
timestamp=$(date +%Y%m%d_%H%M%S)
|
| 268 |
-
backup_file="${BACKUP_PREFIX}_${timestamp}.tar.gz"
|
| 269 |
-
backup_path="/tmp/${backup_file}" # Use /tmp for temporary files
|
| 270 |
-
|
| 271 |
-
echo "Compressing Cloudreve data (executable, db, config) to: $backup_path"
|
| 272 |
-
# Use -C to change directory, ensuring archive paths are relative
|
| 273 |
-
# Only include the executable, db, and config file
|
| 274 |
-
tar -czf "$backup_path" -C "$CLOUDREVE_DIR" \
|
| 275 |
-
$(basename "$EXECUTABLE_PATH") \
|
| 276 |
-
$(basename "$DB_FILE_PATH") \
|
| 277 |
-
$(basename "$CONFIG_FILE_PATH")
|
| 278 |
-
|
| 279 |
-
# Check if compression was successful (file exists and is not empty)
|
| 280 |
-
if [ -s "$backup_path" ]; then
|
| 281 |
-
echo "Compression complete. File size: $(ls -lh "$backup_path" | awk '{print $5}')"
|
| 282 |
-
echo "Uploading backup to HuggingFace..."
|
| 283 |
-
upload_backup "$backup_path" "${backup_file}"
|
| 284 |
-
# Check exit status of upload_backup? The python script should exit non-zero on failure.
|
| 285 |
-
if [ $? -ne 0 ]; then
|
| 286 |
-
echo "Backup upload failed. Keeping local archive: $backup_path"
|
| 287 |
-
else
|
| 288 |
-
echo "Upload successful. Removing local archive."
|
| 289 |
-
rm -f "$backup_path"
|
| 290 |
-
fi
|
| 291 |
-
else
|
| 292 |
-
echo "Compression failed or created an empty file. Skipping upload."
|
| 293 |
-
rm -f "$backup_path" # Remove potentially empty/corrupt file
|
| 294 |
-
fi
|
| 295 |
-
|
| 296 |
-
# Define sync interval (use environment variable or default to 3600 seconds = 1 hour)
|
| 297 |
-
# **重要:确保这里的间隔设置合理,避免过于频繁导致资源超限**
|
| 298 |
-
SYNC_INTERVAL=${SYNC_INTERVAL:-3600} # 默认改为 1 小时
|
| 299 |
-
echo "Next sync in ${SYNC_INTERVAL} seconds..."
|
| 300 |
-
sleep $SYNC_INTERVAL
|
| 301 |
done
|
| 302 |
}
|
| 303 |
|
| 304 |
-
# --- Main Execution ---
|
| 305 |
|
| 306 |
-
#
|
| 307 |
-
|
| 308 |
-
|
| 309 |
-
|
| 310 |
-
|
| 311 |
-
|
| 312 |
-
|
| 313 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
exit 1
|
| 315 |
-
elif [
|
| 316 |
-
echo "
|
|
|
|
|
|
|
|
|
|
|
|
|
| 317 |
fi
|
|
|
|
|
|
|
| 318 |
|
| 319 |
|
| 320 |
-
# 2.
|
|
|
|
| 321 |
if [ ! -f "$CONFIG_FILE_PATH" ]; then
|
| 322 |
-
echo "Config file ($CONFIG_FILE_PATH) not found. Running Cloudreve once to generate initial config
|
| 323 |
-
# 首次运行时,不需要后台运行,让它生成配置就退出
|
| 324 |
/opt/cloudreve/cloudreve -c "$CONFIG_FILE_PATH"
|
| 325 |
-
# Cloudreve 在没有数据库时会初始化并打印密码然后退出
|
| 326 |
if [ ! -f "$CONFIG_FILE_PATH" ]; then
|
| 327 |
echo "CRITICAL: Cloudreve failed to create initial config file. Exiting."
|
| 328 |
exit 1
|
| 329 |
else
|
| 330 |
-
echo "Initial config file created by first run.
|
| 331 |
-
# 确保数据库文件也可能已创建(即使是空的)
|
| 332 |
-
if [ ! -f "$DB_FILE_PATH" ]; then
|
| 333 |
-
echo "WARN: Database file might not have been created on first run, proceeding anyway."
|
| 334 |
-
fi
|
| 335 |
fi
|
| 336 |
fi
|
| 337 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 338 |
|
| 339 |
-
#
|
| 340 |
-
|
| 341 |
-
|
| 342 |
-
|
| 343 |
-
|
| 344 |
-
|
| 345 |
-
#
|
| 346 |
-
#
|
| 347 |
-
|
| 348 |
-
|
| 349 |
-
|
| 350 |
-
|
| 351 |
-
|
| 352 |
-
|
| 353 |
-
echo "
|
| 354 |
-
|
| 355 |
-
|
| 356 |
-
|
|
|
|
|
|
| 1 |
#!/bin/bash
|
| 2 |
|
| 3 |
+
# Function to check if Cloudreve is running by checking the port
|
| 4 |
+
wait_for_cloudreve() {
|
| 5 |
+
echo "Waiting for Cloudreve service to become available on port 5212..."
|
| 6 |
+
# Loop until nc successfully connects (-z checks without sending data)
|
| 7 |
+
while ! nc -z 127.0.0.1 5212; do
|
| 8 |
+
echo "Cloudreve not ready yet (port 5212 not open), sleeping for 5 seconds..."
|
| 9 |
+
sleep 5
|
| 10 |
+
done
|
| 11 |
+
echo "Cloudreve service detected on port 5212."
|
| 12 |
+
}
|
| 13 |
+
|
| 14 |
+
# Function to perform a single backup action
|
| 15 |
+
perform_backup_once() {
|
| 16 |
+
echo "Performing single backup operation..."
|
| 17 |
+
local config_path="$1"
|
| 18 |
+
local db_path="$2"
|
| 19 |
+
local exe_path="$3"
|
| 20 |
+
local cloudreve_dir="$4"
|
| 21 |
+
local backup_prefix="$5"
|
| 22 |
+
|
| 23 |
+
if [ ! -f "$config_path" ] || [ ! -f "$db_path" ] || [ ! -f "$exe_path" ]; then
|
| 24 |
+
echo "WARN: Essential files missing for backup. Skipping this backup."
|
| 25 |
+
return 1
|
| 26 |
+
fi
|
| 27 |
+
|
| 28 |
+
local timestamp=$(date +%Y%m%d_%H%M%S)
|
| 29 |
+
local backup_file="${backup_prefix}_${timestamp}.tar.gz"
|
| 30 |
+
local backup_path="/tmp/${backup_file}" # Use /tmp for temporary files
|
| 31 |
+
|
| 32 |
+
echo "Compressing Cloudreve data (executable, db, config) to: $backup_path"
|
| 33 |
+
tar -czf "$backup_path" -C "$cloudreve_dir" \
|
| 34 |
+
$(basename "$exe_path") \
|
| 35 |
+
$(basename "$db_path") \
|
| 36 |
+
$(basename "$config_path")
|
| 37 |
+
|
| 38 |
+
if [ -s "$backup_path" ]; then
|
| 39 |
+
echo "Compression complete. File size: $(ls -lh "$backup_path" | awk '{print $5}')"
|
| 40 |
+
echo "Uploading backup to HuggingFace..."
|
| 41 |
+
# Call the previously defined upload_backup function
|
| 42 |
+
upload_backup "$backup_path" "${backup_file}"
|
| 43 |
+
local upload_status=$?
|
| 44 |
+
if [ $upload_status -ne 0 ]; then
|
| 45 |
+
echo "Backup upload failed. Keeping local archive: $backup_path"
|
| 46 |
+
return 1
|
| 47 |
+
else
|
| 48 |
+
echo "Upload successful. Removing local archive."
|
| 49 |
+
rm -f "$backup_path"
|
| 50 |
+
return 0
|
| 51 |
+
fi
|
| 52 |
+
else
|
| 53 |
+
echo "Compression failed or created an empty file. Skipping upload."
|
| 54 |
+
rm -f "$backup_path" # Remove potentially empty/corrupt file
|
| 55 |
+
return 1
|
| 56 |
+
fi
|
| 57 |
+
}
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
# --- Start Script Execution ---
|
| 61 |
+
|
| 62 |
+
# Check Hugging Face Token and Dataset ID environment variables
|
| 63 |
if [[ -z "$HF_TOKEN" ]] || [[ -z "$DATASET_ID" ]]; then
|
| 64 |
+
echo "Starting without backup functionality - missing HF_TOKEN or DATASET_ID"
|
| 65 |
+
# Ensure Aria2 starts if needed, even without backup/restore
|
| 66 |
+
# Example (adjust options as needed from your Dockerfile CMD):
|
| 67 |
+
# aria2c --daemon=true [OTHER_OPTIONS] &
|
|
|
|
|
|
|
| 68 |
exec /opt/cloudreve/cloudreve -c /opt/cloudreve/config.ini
|
| 69 |
+
exit 0
|
| 70 |
fi
|
| 71 |
|
| 72 |
+
# Activate Python virtual environment
|
| 73 |
echo "Activating Python venv..."
|
| 74 |
source /opt/venv/bin/activate
|
| 75 |
|
| 76 |
+
# Define Cloudreve paths and backup prefix
|
| 77 |
CLOUDREVE_DIR="/opt/cloudreve"
|
| 78 |
BACKUP_PREFIX="cloudreve_backup"
|
| 79 |
CONFIG_FILE_PATH="/opt/cloudreve/config.ini"
|
| 80 |
DB_FILE_PATH="/opt/cloudreve/cloudreve.db"
|
| 81 |
EXECUTABLE_PATH="/opt/cloudreve/cloudreve"
|
| 82 |
|
|
|
|
| 83 |
|
| 84 |
+
# --- Python Function Definitions (Ensure these are complete) ---
|
| 85 |
+
|
| 86 |
+
# Python Function: Upload Backup
|
| 87 |
upload_backup() {
|
| 88 |
+
local file_path="$1"
|
| 89 |
+
local file_name="$2"
|
| 90 |
+
local token="$HF_TOKEN"
|
| 91 |
+
local repo_id="$DATASET_ID"
|
| 92 |
|
| 93 |
echo "Preparing to upload backup file: $file_path as $file_name to Dataset: $repo_id"
|
| 94 |
|
| 95 |
+
# Use python3 -c "..." with the full Python code block here
|
| 96 |
+
# (Using placeholder for brevity, replace with your actual Python code)
|
| 97 |
python3 -c "
|
| 98 |
from huggingface_hub import HfApi
|
| 99 |
+
import sys, os
|
|
|
|
| 100 |
print(f'HF_TOKEN is set: {os.environ.get(\"HF_TOKEN\") is not None}')
|
| 101 |
print(f'DATASET_ID is set: {os.environ.get(\"DATASET_ID\") is not None}')
|
| 102 |
+
def manage_backups(api, repo_id_val, max_files=50): # Increased max_files
|
| 103 |
print('Managing old backups...')
|
| 104 |
+
try:
|
| 105 |
+
files = api.list_repo_files(repo_id=repo_id_val, repo_type='dataset')
|
| 106 |
+
backup_files = sorted([f for f in files if f.startswith('$BACKUP_PREFIX') and f.endswith('.tar.gz')])
|
| 107 |
+
if len(backup_files) >= max_files:
|
| 108 |
+
print(f'Found {len(backup_files)} backup files, maximum allowed is {max_files}.')
|
| 109 |
+
files_to_delete = backup_files[:len(backup_files) - max_files + 1]
|
| 110 |
+
print(f'Will delete {len(files_to_delete)} old backups.')
|
| 111 |
+
for file_to_delete in files_to_delete:
|
| 112 |
+
try:
|
| 113 |
+
print(f'Deleting old backup: {file_to_delete}')
|
| 114 |
+
api.delete_file(path_in_repo=file_to_delete, repo_id=repo_id_val, repo_type='dataset', token=os.environ.get('HF_TOKEN'))
|
| 115 |
+
print(f'Successfully deleted: {file_to_delete}')
|
| 116 |
+
except Exception as e:
|
| 117 |
+
print(f'Error deleting {file_to_delete}: {str(e)}')
|
| 118 |
+
else:
|
| 119 |
+
print(f'Number of backup files ({len(backup_files)}) is within the limit ({max_files}).')
|
| 120 |
+
except Exception as e:
|
| 121 |
+
print(f'Error listing or managing backups: {e}')
|
| 122 |
+
|
| 123 |
api = HfApi(token='$token')
|
| 124 |
try:
|
| 125 |
+
repo_id_val = os.environ.get('DATASET_ID')
|
| 126 |
+
if not repo_id_val: raise ValueError('DATASET_ID environment variable is not set.')
|
|
|
|
| 127 |
print(f'Uploading file: $file_path to {repo_id_val} as $file_name')
|
| 128 |
+
api.upload_file(path_or_fileobj='$file_path', path_in_repo='$file_name', repo_id=repo_id_val, repo_type='dataset', token=os.environ.get('HF_TOKEN'))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
print(f'Successfully uploaded $file_name')
|
| 130 |
manage_backups(api, repo_id_val)
|
| 131 |
except Exception as e:
|
| 132 |
print(f'Error uploading file: {str(e)}')
|
| 133 |
+
sys.exit(1) # Indicate failure
|
| 134 |
"
|
| 135 |
+
# Capture python exit code
|
| 136 |
+
return $?
|
| 137 |
}
|
| 138 |
|
| 139 |
+
# Python Function: Download Latest Backup
|
| 140 |
download_latest_backup() {
|
| 141 |
+
local token="$HF_TOKEN"
|
| 142 |
+
local repo_id="$DATASET_ID"
|
| 143 |
+
local download_flag_file="/tmp/backup_restored.flag" # Flag file
|
| 144 |
|
| 145 |
+
# Remove any previous flag file
|
| 146 |
+
rm -f "$download_flag_file"
|
| 147 |
|
| 148 |
+
echo "Preparing to download the latest backup from Dataset: $repo_id"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 149 |
|
| 150 |
+
# Use python3 -c "..." with the full Python code block here
|
| 151 |
+
# (Using placeholder for brevity, replace with your actual Python code)
|
| 152 |
+
python3 -c "
|
| 153 |
+
from huggingface_hub import HfApi, hf_hub_download
|
| 154 |
+
import sys, os, tarfile, tempfile, shutil, subprocess
|
| 155 |
print(f'HF_TOKEN is set: {os.environ.get(\"HF_TOKEN\") is not None}')
|
| 156 |
print(f'DATASET_ID is set: {os.environ.get(\"DATASET_ID\") is not None}')
|
| 157 |
+
flag_file = '$download_flag_file' # Get flag file path
|
| 158 |
|
| 159 |
api = HfApi(token='$token')
|
| 160 |
try:
|
| 161 |
+
repo_id_val = os.environ.get('DATASET_ID')
|
| 162 |
+
if not repo_id_val: raise ValueError('DATASET_ID environment variable is not set.')
|
|
|
|
| 163 |
|
| 164 |
print(f'Listing files in Dataset: {repo_id_val}')
|
| 165 |
files = api.list_repo_files(repo_id=repo_id_val, repo_type='dataset')
|
| 166 |
+
backup_files = sorted([f for f in files if f.startswith('$BACKUP_PREFIX') and f.endswith('.tar.gz')])
|
| 167 |
|
| 168 |
if not backup_files:
|
| 169 |
print('No backup files found in the Dataset. Skipping restore.')
|
| 170 |
+
sys.exit(0) # Success: No backup found is not an error for the script flow
|
|
|
|
| 171 |
|
| 172 |
+
latest_backup = backup_files[-1]
|
| 173 |
print(f'Latest backup file found: {latest_backup}')
|
| 174 |
|
| 175 |
with tempfile.TemporaryDirectory() as temp_dir:
|
| 176 |
print(f'Downloading {latest_backup} to temporary directory {temp_dir}...')
|
| 177 |
try:
|
| 178 |
+
filepath = hf_hub_download(repo_id=repo_id_val, filename=latest_backup, repo_type='dataset', local_dir=temp_dir, token=os.environ.get('HF_TOKEN'))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 179 |
except Exception as download_error:
|
| 180 |
+
print(f'Error during hf_hub_download: {download_error}'); sys.exit(1)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 181 |
|
| 182 |
if filepath and os.path.exists(filepath):
|
| 183 |
print(f'Successfully downloaded backup to temporary directory: {filepath}')
|
|
|
|
|
|
|
| 184 |
items_to_restore = ['cloudreve', 'cloudreve.db', 'config.ini']
|
|
|
|
|
|
|
| 185 |
os.makedirs(\"$CLOUDREVE_DIR\", exist_ok=True)
|
| 186 |
+
print('Listing contents before restore:'); subprocess.run(['ls', '-lA', \"$CLOUDREVE_DIR\"], check=False)
|
| 187 |
|
| 188 |
+
extract_temp_dir = os.path.join(temp_dir, 'extracted_backup'); os.makedirs(extract_temp_dir, exist_ok=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 189 |
print(f'Extracting backup archive: {filepath} to {extract_temp_dir}')
|
| 190 |
try:
|
| 191 |
+
with tarfile.open(filepath, 'r:gz') as tar: tar.extractall(extract_temp_dir)
|
|
|
|
|
|
|
| 192 |
print('Extraction complete.')
|
| 193 |
+
except Exception as extract_err: print(f'Error during extraction: {extract_err}'); sys.exit(1)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 194 |
|
|
|
|
| 195 |
essential_files_present = True
|
|
|
|
| 196 |
for item in items_to_restore:
|
| 197 |
+
extracted_item_path = os.path.join(extract_temp_dir, item)
|
| 198 |
+
if not os.path.exists(extracted_item_path):
|
| 199 |
+
print(f'Error: Essential item "{item}" not found in extracted backup at {extracted_item_path}. Aborting restore.')
|
| 200 |
+
essential_files_present = False; # break # Keep checking all missing items
|
|
|
|
|
|
|
|
|
|
| 201 |
if not essential_files_present:
|
| 202 |
+
print('Listing contents of extraction directory for debugging:'); subprocess.run(['ls', '-lA', extract_temp_dir], check=False)
|
|
|
|
| 203 |
sys.exit(1) # Abort if essential files are missing
|
| 204 |
|
|
|
|
| 205 |
print(f'Deleting existing items in $CLOUDREVE_DIR before restoring...')
|
| 206 |
for item in items_to_restore:
|
| 207 |
target_path = os.path.join(\"$CLOUDREVE_DIR\", item)
|
| 208 |
if os.path.exists(target_path):
|
| 209 |
try:
|
| 210 |
+
if os.path.isdir(target_path) and not os.path.islink(target_path): shutil.rmtree(target_path)
|
| 211 |
+
else: os.remove(target_path)
|
| 212 |
+
except OSError as e: print(f'Error deleting {target_path}: {e}. Continuing...')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 213 |
|
|
|
|
| 214 |
print(f'Moving extracted items from {extract_temp_dir} to $CLOUDREVE_DIR...')
|
| 215 |
for item in items_to_restore:
|
| 216 |
source_path = os.path.join(extract_temp_dir, item)
|
|
|
|
| 217 |
if os.path.exists(source_path):
|
| 218 |
target_path = os.path.join(\"$CLOUDREVE_DIR\", item)
|
| 219 |
+
try: shutil.move(source_path, target_path)
|
| 220 |
+
except Exception as move_err: print(f'Error moving {item}: {move_err}') # Continue for now
|
| 221 |
+
else: print(f'Warning: Source item {item} disappeared before move.')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 222 |
|
| 223 |
print(f'Successfully restored backup from {latest_backup}')
|
| 224 |
+
print('Listing contents after restore:'); subprocess.run(['ls', '-lA', \"$CLOUDREVE_DIR\"], check=False)
|
| 225 |
+
# Create the flag file only on successful restore
|
| 226 |
+
with open(flag_file, 'w') as f: f.write('restored')
|
| 227 |
+
print(f'Created restore flag file: {flag_file}')
|
| 228 |
+
sys.exit(0) # Indicate successful restore completion
|
| 229 |
else:
|
| 230 |
print(f'Error: Downloaded file path "{filepath}" does not exist or download failed.')
|
| 231 |
+
sys.exit(1)
|
| 232 |
|
| 233 |
+
except ValueError as ve: print(f'Configuration Error: {ve}'); sys.exit(1)
|
| 234 |
+
except Exception as e: import traceback; print(f'Error during backup download/restore: {str(e)}'); traceback.print_exc(); sys.exit(1)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 235 |
"
|
| 236 |
+
# Capture python exit code
|
| 237 |
+
return $?
|
| 238 |
}
|
| 239 |
|
| 240 |
+
# --- Periodic Sync Function (Loop only) ---
|
| 241 |
+
sync_data_loop() {
|
| 242 |
+
echo "Starting Periodic Background Sync Loop..."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
while true; do
|
| 244 |
+
# Wait for the defined interval
|
| 245 |
+
local sync_interval=${SYNC_INTERVAL:-3600} # Default to 1 hour
|
| 246 |
+
echo "Periodic Sync: Next check in ${sync_interval} seconds..."
|
| 247 |
+
sleep $sync_interval
|
| 248 |
+
|
| 249 |
+
# Perform the backup
|
| 250 |
+
echo "Periodic Sync: Starting backup cycle at $(date)"
|
| 251 |
+
perform_backup_once "$CONFIG_FILE_PATH" "$DB_FILE_PATH" "$EXECUTABLE_PATH" "$CLOUDREVE_DIR" "$BACKUP_PREFIX"
|
| 252 |
+
# We don't necessarily need to exit the loop if one backup fails
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 253 |
done
|
| 254 |
}
|
| 255 |
|
|
|
|
| 256 |
|
| 257 |
+
# --- Main Execution Logic ---
|
| 258 |
+
|
| 259 |
+
# 1. Attempt Initial Restore
|
| 260 |
+
echo "--- Step 1: Attempting Initial Restore ---"
|
| 261 |
+
download_latest_backup
|
| 262 |
+
restore_exit_code=$?
|
| 263 |
+
restore_flag_file="/tmp/backup_restored.flag"
|
| 264 |
+
|
| 265 |
+
# Check restore outcome
|
| 266 |
+
backup_was_restored=false
|
| 267 |
+
if [ $restore_exit_code -ne 0 ]; then
|
| 268 |
+
echo "CRITICAL: Backup restoration attempt failed with error (Exit code: $restore_exit_code). Exiting."
|
| 269 |
exit 1
|
| 270 |
+
elif [ -f "$restore_flag_file" ]; then
|
| 271 |
+
echo "Restore successful: Backup found and restored."
|
| 272 |
+
backup_was_restored=true
|
| 273 |
+
else
|
| 274 |
+
echo "Restore skipped: No backup file found in the dataset."
|
| 275 |
+
backup_was_restored=false
|
| 276 |
fi
|
| 277 |
+
# Clean up flag file regardless
|
| 278 |
+
rm -f "$restore_flag_file"
|
| 279 |
|
| 280 |
|
| 281 |
+
# 2. Initial Cloudreve Run (If necessary)
|
| 282 |
+
echo "--- Step 2: Checking for Initial Config ---"
|
| 283 |
if [ ! -f "$CONFIG_FILE_PATH" ]; then
|
| 284 |
+
echo "Config file ($CONFIG_FILE_PATH) not found. Running Cloudreve once to generate initial config..."
|
|
|
|
| 285 |
/opt/cloudreve/cloudreve -c "$CONFIG_FILE_PATH"
|
|
|
|
| 286 |
if [ ! -f "$CONFIG_FILE_PATH" ]; then
|
| 287 |
echo "CRITICAL: Cloudreve failed to create initial config file. Exiting."
|
| 288 |
exit 1
|
| 289 |
else
|
| 290 |
+
echo "Initial config file created by first run. Check logs for credentials if needed."
|
|
|
|
|
|
|
|
|
|
|
|
|
| 291 |
fi
|
| 292 |
fi
|
| 293 |
|
| 294 |
+
# 3. Start Cloudreve Main Process (in background temporarily to allow checks)
|
| 295 |
+
echo "--- Step 3: Starting Cloudreve Main Process (background initially) ---"
|
| 296 |
+
/opt/cloudreve/cloudreve -c "$CONFIG_FILE_PATH" &
|
| 297 |
+
cloudreve_pid=$!
|
| 298 |
+
|
| 299 |
+
# 4. Wait for Cloudreve to be Ready
|
| 300 |
+
echo "--- Step 4: Waiting for Cloudreve Service ---"
|
| 301 |
+
wait_for_cloudreve # Use the function defined earlier
|
| 302 |
+
|
| 303 |
+
# 5. Perform Initial Backup (If No Restore Occurred)
|
| 304 |
+
echo "--- Step 5: Checking for Initial Backup ---"
|
| 305 |
+
if [ "$backup_was_restored" = false ]; then
|
| 306 |
+
echo "No backup was restored, performing initial backup of current state..."
|
| 307 |
+
perform_backup_once "$CONFIG_FILE_PATH" "$DB_FILE_PATH" "$EXECUTABLE_PATH" "$CLOUDREVE_DIR" "$BACKUP_PREFIX"
|
| 308 |
+
if [ $? -ne 0 ]; then
|
| 309 |
+
echo "Warning: Initial backup failed."
|
| 310 |
+
# Decide whether to continue or exit if initial backup fails
|
| 311 |
+
else
|
| 312 |
+
echo "Initial backup completed successfully."
|
| 313 |
+
fi
|
| 314 |
+
else
|
| 315 |
+
echo "Backup was restored, skipping initial backup."
|
| 316 |
+
fi
|
| 317 |
|
| 318 |
+
# 6. Start Periodic Sync Loop in Background
|
| 319 |
+
echo "--- Step 6: Starting Periodic Sync Loop ---"
|
| 320 |
+
sync_data_loop &
|
| 321 |
+
sync_loop_pid=$!
|
| 322 |
+
|
| 323 |
+
# 7. Bring Cloudreve to Foreground / Wait for it
|
| 324 |
+
# Since Cloudreve was started in the background, we now wait for it.
|
| 325 |
+
# If Cloudreve exits, this script will also exit.
|
| 326 |
+
echo "--- Step 7: Cloudreve is running. Monitoring PID: $cloudreve_pid ---"
|
| 327 |
+
wait $cloudreve_pid
|
| 328 |
+
exit_code=$?
|
| 329 |
+
echo "Cloudreve process exited with code $exit_code."
|
| 330 |
+
|
| 331 |
+
# Optional: attempt to clean up background sync loop if Cloudreve exits unexpectedly
|
| 332 |
+
echo "Attempting to terminate background sync loop (PID: $sync_loop_pid)..."
|
| 333 |
+
kill $sync_loop_pid 2>/dev/null
|
| 334 |
+
wait $sync_loop_pid 2>/dev/null # Wait briefly for it to terminate
|
| 335 |
+
|
| 336 |
+
exit $exit_code
|