feat(device): 无障碍保护存活心跳检测 + 掉线终端告警
- 新增 device 表 + /api/v1/device/{register,heartbeat} + 迁移 device_table
- heartbeat_monitor_worker 周期扫描心跳超时(App 被杀/无障碍停)→ 服务器终端打印告警
(推送本期未接,先用 logger 终端打印代替真实通知;integrations/jpush.py 已备,后续直接替换)
- config / .env.example 增 JPUSH_* / HEARTBEAT_*
- 见 spec(仓库外 e:\codes\spec\accessibility-liveness-push.md)
注:本提交同时快照了并行 session 未提交的 coupon_state 改动 + 相关 migration
(与本功能同迁移链耦合,无法单独拆分,经确认一并提交)。
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
陈世睿
@@ -61,6 +61,32 @@ class Settings(BaseSettings):
|
||||
JG_VERIFY_ENDPOINT: str = "https://api.verification.jpush.cn/v1/web/loginTokenVerify"
|
||||
JG_REQUEST_TIMEOUT_SEC: int = 15
|
||||
|
||||
# ===== 极光推送 JPush(无障碍保护掉线告警)=====
|
||||
# 客户端 push appkey 已知 = 966b451a8d9cfe12d173ea9d(android build.gradle manifestPlaceholder)。
|
||||
# 若推送与一键登录/短信是同一个极光应用(大概率),JPUSH_* 留空即自动回退到 JG_*。
|
||||
# 否则在 .env 单独配 JPUSH_APP_KEY / JPUSH_MASTER_SECRET(对应那个 push appkey)。
|
||||
JPUSH_APP_KEY: str = ""
|
||||
JPUSH_MASTER_SECRET: str = ""
|
||||
JPUSH_PUSH_ENDPOINT: str = "https://api.jpush.cn/v3/push"
|
||||
|
||||
# 无障碍保护存活监控后台任务
|
||||
HEARTBEAT_MONITOR_ENABLED: bool = True # 总开关
|
||||
HEARTBEAT_TIMEOUT_MINUTES: int = 10 # 多久没心跳算掉线(≈3 个客户端心跳周期)
|
||||
HEARTBEAT_SCAN_INTERVAL_SEC: int = 60 # 扫描周期
|
||||
|
||||
@property
|
||||
def jpush_app_key(self) -> str:
|
||||
return self.JPUSH_APP_KEY or self.JG_APP_KEY
|
||||
|
||||
@property
|
||||
def jpush_master_secret(self) -> str:
|
||||
return self.JPUSH_MASTER_SECRET or self.JG_MASTER_SECRET
|
||||
|
||||
@property
|
||||
def jpush_configured(self) -> bool:
|
||||
"""推送凭证齐全(缺则 monitor 只扫不发,不报错)。"""
|
||||
return bool(self.jpush_app_key and self.jpush_master_secret)
|
||||
|
||||
# ===== 短信 =====
|
||||
SMS_MOCK: bool = True
|
||||
SMS_CODE_TTL_SEC: int = 300
|
||||
|
||||
@@ -0,0 +1,148 @@
|
||||
"""无障碍保护存活监控后台任务。
|
||||
|
||||
周期扫描「曾经保护过、当前 alive、心跳超时」的设备 = App 被彻底杀掉/无障碍已停(心跳断了),
|
||||
**命中即在服务器终端打印告警**(本期先不接推送,工程量大,用终端打印代替真实通知);并把状态机
|
||||
推进到 notified 防每轮重复打印(心跳恢复时由 repositories.device.touch_heartbeat 重置回 alive)。
|
||||
结构仿 withdraw_reconcile_worker(单实例锁 + asyncio 轮询 + 优雅退出)。
|
||||
|
||||
见 spec: spec/accessibility-liveness-push.md。
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import asyncio
|
||||
import contextlib
|
||||
import logging
|
||||
import os
|
||||
import time
|
||||
from collections.abc import Iterator
|
||||
from datetime import datetime, timezone
|
||||
from pathlib import Path
|
||||
|
||||
from sqlalchemy.exc import SQLAlchemyError
|
||||
|
||||
from app.core.config import settings
|
||||
from app.db.session import SessionLocal
|
||||
from app.repositories import device as device_repo
|
||||
|
||||
logger = logging.getLogger("shagua.heartbeat_monitor")
|
||||
_LOCK_PATH = Path(__file__).resolve().parents[2] / "data" / "heartbeat_monitor.lock"
|
||||
|
||||
|
||||
def _touch_lock() -> None:
|
||||
with contextlib.suppress(FileNotFoundError):
|
||||
os.utime(_LOCK_PATH, None)
|
||||
|
||||
|
||||
@contextlib.contextmanager
|
||||
def _single_instance_lock(stale_after_sec: int) -> Iterator[bool]:
|
||||
"""同机多进程保护:同一时间只允许一个监控 worker 运行。"""
|
||||
_LOCK_PATH.parent.mkdir(parents=True, exist_ok=True)
|
||||
fd: int | None = None
|
||||
try:
|
||||
try:
|
||||
fd = os.open(str(_LOCK_PATH), os.O_CREAT | os.O_EXCL | os.O_WRONLY)
|
||||
except FileExistsError:
|
||||
try:
|
||||
age = time.time() - _LOCK_PATH.stat().st_mtime
|
||||
except FileNotFoundError:
|
||||
age = stale_after_sec + 1
|
||||
if age > stale_after_sec:
|
||||
with contextlib.suppress(FileNotFoundError):
|
||||
_LOCK_PATH.unlink()
|
||||
try:
|
||||
fd = os.open(str(_LOCK_PATH), os.O_CREAT | os.O_EXCL | os.O_WRONLY)
|
||||
except FileExistsError:
|
||||
fd = None
|
||||
|
||||
if fd is None:
|
||||
yield False
|
||||
return
|
||||
|
||||
os.write(fd, f"pid={os.getpid()} started_at={int(time.time())}\n".encode("ascii"))
|
||||
yield True
|
||||
finally:
|
||||
if fd is not None:
|
||||
os.close(fd)
|
||||
with contextlib.suppress(FileNotFoundError):
|
||||
_LOCK_PATH.unlink()
|
||||
|
||||
|
||||
def _silent_seconds(last: datetime | None) -> int | None:
|
||||
"""距上次心跳的秒数(兼容 sqlite 取回的 naive datetime)。"""
|
||||
if last is None:
|
||||
return None
|
||||
ref = datetime.now(timezone.utc) if last.tzinfo is not None else datetime.utcnow()
|
||||
return int((ref - last).total_seconds())
|
||||
|
||||
|
||||
def _scan_once(timeout_minutes: int) -> dict:
|
||||
"""扫描一轮:找出心跳超时(App 被彻底杀掉/无障碍已停)的设备,在**服务器终端打印**告警代替真实推送。
|
||||
|
||||
本期不接推送(极光/厂商通道工程量大),只做服务端掉线检测:命中即 logger.warning 打印到终端,
|
||||
并把状态机推进到 notified 防每轮重复打印(心跳恢复时 touch_heartbeat 会重置回 alive)。
|
||||
"""
|
||||
notified = 0
|
||||
with SessionLocal() as db:
|
||||
overdue = device_repo.list_overdue(db, timeout_minutes=timeout_minutes)
|
||||
for device in overdue:
|
||||
silent = _silent_seconds(device.last_heartbeat_at)
|
||||
logger.warning(
|
||||
"🔴 [掉线检测] user_id=%s device_id=%s 已 %s 秒无心跳(阈值 %d 分钟)"
|
||||
" → 判定 App 已被杀/无障碍已停。【本应推送通知提醒用户重开;推送暂未接,先终端打印代替】",
|
||||
device.user_id,
|
||||
device.device_id,
|
||||
silent if silent is not None else "?",
|
||||
timeout_minutes,
|
||||
)
|
||||
device_repo.mark_notified(db, device_id_pk=device.id)
|
||||
notified += 1
|
||||
return {"checked": len(overdue), "notified": notified}
|
||||
|
||||
|
||||
async def _run_loop() -> None:
|
||||
interval = max(10, int(settings.HEARTBEAT_SCAN_INTERVAL_SEC))
|
||||
timeout_minutes = max(1, int(settings.HEARTBEAT_TIMEOUT_MINUTES))
|
||||
lock_stale_after = max(interval * 3, 600)
|
||||
with _single_instance_lock(lock_stale_after) as lock_acquired:
|
||||
if not lock_acquired:
|
||||
logger.warning("heartbeat monitor skipped: another worker owns lock")
|
||||
return
|
||||
await _run_locked_loop(interval, timeout_minutes)
|
||||
|
||||
|
||||
async def _run_locked_loop(interval: int, timeout_minutes: int) -> None:
|
||||
logger.info(
|
||||
"heartbeat monitor started interval=%ss timeout=%sm",
|
||||
interval,
|
||||
timeout_minutes,
|
||||
)
|
||||
try:
|
||||
while True:
|
||||
try:
|
||||
_touch_lock()
|
||||
result = await asyncio.to_thread(_scan_once, timeout_minutes)
|
||||
if result["notified"]:
|
||||
logger.info("heartbeat monitor result=%s", result)
|
||||
except SQLAlchemyError:
|
||||
logger.exception("heartbeat monitor db error")
|
||||
except Exception: # noqa: BLE001 - 后台任务不能因单次异常退出
|
||||
logger.exception("heartbeat monitor unexpected error")
|
||||
await asyncio.sleep(interval)
|
||||
except asyncio.CancelledError:
|
||||
logger.info("heartbeat monitor stopped")
|
||||
raise
|
||||
|
||||
|
||||
def start_heartbeat_monitor() -> asyncio.Task | None:
|
||||
if not settings.HEARTBEAT_MONITOR_ENABLED:
|
||||
logger.info("heartbeat monitor disabled")
|
||||
return None
|
||||
return asyncio.create_task(_run_loop(), name="heartbeat-monitor")
|
||||
|
||||
|
||||
async def stop_heartbeat_monitor(task: asyncio.Task | None) -> None:
|
||||
if task is None:
|
||||
return
|
||||
task.cancel()
|
||||
with contextlib.suppress(asyncio.CancelledError):
|
||||
await task
|
||||
Reference in New Issue
Block a user