depclau / sync_data.sh
nbugs's picture
Update sync_data.sh
4c3b9c0 verified
raw
history blame
2.48 kB
#!/bin/bash
if [ -z "$HF_TOKEN" ] || [ -z "$DATASET_ID" ]; then
echo "Starting without backup - missing HF_TOKEN or DATASET_ID"
exec uvicorn app.main:app --host 0.0.0.0 --port 7860
exit 0
fi
export HUGGING_FACE_HUB_TOKEN=$HF_TOKEN
restore_latest() {
echo "正在检查是否有可用备份..."
python3 -c "
from huggingface_hub import HfApi
import os
api = HfApi()
files = api.list_repo_files('${DATASET_ID}', repo_type='dataset')
backup_files = sorted([f for f in files if f.startswith('backup_') and f.endswith('.tar.gz')])
if backup_files:
latest = backup_files[-1]
print(f'发现备份文件: {latest}, 正在下载...')
api.hf_hub_download(repo_id='${DATASET_ID}', filename=latest, repo_type='dataset', local_dir='/tmp')
# 解压文件并禁止恢复原始权限和所有者,避免容器内权限错误
result = os.system(f'tar --no-same-owner --no-same-permissions -xzf /tmp/{latest} -C /app')
if result == 0:
print(f'成功从 {latest} 恢复备份数据')
else:
print(f'解压备份文件 {latest} 时出现问题,请检查相关日志')
os.remove(f'/tmp/{latest}')
else:
print('未找到备份文件,已跳过数据恢复步骤')
"
}
sync_data() {
echo "Waiting ${SYNC_INTERVAL:-7200} seconds before first backup..."
sleep ${SYNC_INTERVAL:-7200}
while true; do
echo "Starting backup at $(date)"
cd /app
timestamp=$(date +%Y%m%d_%H%M%S)
backup_file="backup_${timestamp}.tar.gz"
if [ -d "app" ] && [ "$(ls -A app 2>/dev/null)" ]; then
tar -czf "/tmp/${backup_file}" app/
python3 -c "
from huggingface_hub import HfApi
api = HfApi()
api.upload_file('/tmp/${backup_file}', '${backup_file}', '${DATASET_ID}', repo_type='dataset')
print('Backup uploaded successfully')
backup_files = sorted([f for f in api.list_repo_files('${DATASET_ID}', repo_type='dataset') if f.startswith('backup_')])
for old_backup in backup_files[:-50]:
api.delete_file(path_in_repo=old_backup, repo_id='${DATASET_ID}', repo_type='dataset')
print(f'Deleted old backup: {old_backup}')
"
rm -f "/tmp/${backup_file}"
else
echo "No data to backup"
fi
echo "Next backup in ${SYNC_INTERVAL:-7200} seconds..."
sleep ${SYNC_INTERVAL:-7200}
done
}
(
restore_latest
sync_data &
exec uvicorn app.main:app --host 0.0.0.0 --port 7860
) 2>&1 | tee -a /app/data/backup.log