feat: 接入企微智能机器人——异常报警/数据核验日报/设备健康日报

- flows/shared/wecom.py: WSClient 按需连接(连接→认证→发送→断开),全容错,超长分段
- flows/monitor/alert_failure.py: Prefect 事件触发(Failed/Crashed→run deployment),发群报警
- flows/monitor/daily_report.py: 每日 18:30 近7天 RIIN/SDS 核验(复用 verify)+商品完整性+OCR+失败回顾
- flows/monitor/device_check.py: 每日 09:00 内存/磁盘/GPU(80%阈值)+Docker 容器异常+服务探活
- db.py: 新增 get_goods_integrity/ping; compose worker 挂载 docker.sock+/:/host:ro+GPU
- scripts/wecom_capture_chatid.py: 一次性捕获群 chat_id 工具
This commit is contained in:
yeuimu
2026-08-27 09:40:53 +00:00
parent b9e0bc42be
commit 1d0da2a00b
12 changed files with 836 additions and 0 deletions
+12
View File
@@ -20,6 +20,18 @@ services:
TZ: Asia/Shanghai
depends_on:
- prefect-server
volumes:
# 设备监控(device-check flow):宿主机磁盘使用率 + Docker 容器状态
- /var/run/docker.sock:/var/run/docker.sock
- /:/host:ro
deploy:
resources:
reservations:
devices:
# GPU 监控(nvidia-smi);依赖宿主机 nvidia-container-toolkit
- driver: nvidia
count: all
capabilities: [gpu]
restart: unless-stopped
volumes: