Files

T

Nexus Agent 5c508ed14c fix(watch): eager probe on resume so metrics appear immediately

After pausing clears Redis snapshots, resuming or adding pins now triggers
an immediate probe before list_slots returns, with frontend refresh fallback.

Co-authored-by: Cursor <cursoragent@cursor.com>

2026-06-14 02:53:02 +08:00

1.4 KiB

Raw Blame History

审计 — 监测开启后立即探针

Changelog: docs/changelog/2026-06-14-watch-eager-probe-on-resume.md

审计范围

文件	变更	状态
`watch_probe_runner.py`	抽取 `_probe_server_with_pins`；`trigger_immediate_watch_probe`	☑
`watch_service.py`	加槽/替换/恢复监测后 eager probe	☑
`useWatchPins.ts`	恢复监测 3s/8s 兜底 refresh	☑
`WatchSlotCard.vue`	暂停底栏「已暂停」文案	☑
`test_watch_pins.py`	恢复监测立即 metrics 用例 + noop mock	☑

Step 3 规则扫描

H	规则	结论
H1	无静默吞错	SAFE — `trigger_immediate_watch_probe` 异常 `logger.exception`
H2	探针重复/竞态	SAFE — 与 5s 循环共用 `_probe_server_with_pins`，仅多一次即时采样
H3	暂停清空 Redis	SAFE — 仍仅在 `monitoring_enabled=false` 时 `clear_watch_redis_snapshot`
H4	WS 推送	SAFE — 即时探针同样 `publish_watch_metrics`
H5	前端轮询	SAFE — 仅 `isWatchMetricsPending` 时 refresh，不刷屏

Closure

H	判定	依据
H1–H5	SAFE	恢复监测空窗期根因已消除；测试 10/10；local_verify 26/26

DoD

changelog ≥10 行
pytest tests/test_watch_pins.py
bash scripts/local_verify.sh
生产：开启实时监测后数秒内见指标或探针错误态

1.4 KiB Raw Blame History Unescape Escape

审计 — 监测开启后立即探针

审计范围

Step 3 规则扫描

Closure

DoD

1.4 KiB

Raw Blame History