feat(wave5-p2): GovernanceAgent 4 項自檢 + Ollama 健康告警規則 + Prometheus metrics 整合
All checks were successful
CD Pipeline / build-and-deploy (push) Successful in 10m45s

MASTER plan_complete_v3.md Wave 5 P2.2 + P2.3 完成(multiple engineers 在限額前完成代碼,補 commit):

P2.2 — GovernanceAgent 4 項自檢:
- governance_agent.py (342 行) — 每 1 小時自檢循環:
  · trust_drift(信任度漂移檢測)
  · knowledge_degradation(知識退化檢測)
  · llm_hallucination(LLM 幻覺檢測)
  · execution_blast_radius(執行爆炸半徑檢測)
- main.py lifespan: asyncio.create_task(run_governance_loop()) 啟動
  try/except 包裹,schedule 失敗不阻斷主流程
- failover_alerter.py: alert_governance(event_type, payload) 1h dedup
  四類事件 → Telegram MarkdownV2 告警

P2.3 — Ollama 健康規則 + Prometheus Metrics:
- ops/monitoring/ollama_health_rules.yaml (148 行):
  · OllamaHealthDegraded / OllamaPrimaryDown
  · OllamaFailoverTriggered / GeminiQuotaExceeded
  · 補 Prometheus 取資料的 alert rules
- core/metrics.py (57 行):
  · GEMINI_DAILY_CALL_COUNT / GEMINI_DAILY_QUOTA Gauge
  · OLLAMA_FAILOVER_TRIGGERED_TOTAL Counter
  · OLLAMA_CURRENT_PRIMARY_IS_OLLAMA Gauge
- ollama_failover_manager.py:
  · _check_gemini_quota: 每次 check 同步更新 Gauge(讓 Prometheus 取最新值)
  · select_provider: failover 時 inc Counter + 切 Primary Gauge
  · try/except 包裹,metric 失敗不阻斷主路由

E2E 測試:
- test_failover_e2e_dispatch.py (365 行)
  完整 dispatch 路徑:health check → failover decide → alerter → metrics

Tests: 54 passed (e2e_dispatch + failover_manager + failover_alerter)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-Authored-By: Multiple Engineers (上 session Wave 5) <noreply@anthropic.com>
This commit is contained in:
Your Name
2026-04-26 20:56:19 +08:00
parent bddf99a002
commit 2c57b71db9
7 changed files with 970 additions and 0 deletions

View File

@@ -84,6 +84,32 @@ class FailoverAlerter:
await self._send(msg)
logger.info("recovery_alert_sent", from_provider=from_provider)
async def alert_governance(self, event_type: str, payload: dict[str, Any]) -> None:
"""AI 治理告警dedup 1h
event_type: trust_drift / knowledge_degradation / llm_hallucination / execution_blast_radius
dedup TTL 3600s — 同類告警 1 小時內不重複發送
2026-04-26 P2.2 by Claude
"""
dedup_key = f"alert:governance:{event_type}"
if not await self._check_dedup(dedup_key, ttl=3600):
logger.debug("governance_alert_dedup_skipped", event_type=event_type)
return
# 格式化 payload 為可讀字串key=value換行分隔
detail_lines = "\n".join(
f"{_escape_md(str(k))}{_escape_md(str(v))}"
for k, v in payload.items()
)
msg = (
f"*AI 治理警報*\n\n"
f"類型:{_escape_md(event_type)}\n\n"
f"{detail_lines}"
)
await self._send(msg)
logger.info("governance_alert_sent", event_type=event_type)
async def alert_gemini_quota_exceeded(self, event: dict[str, Any]) -> None:
"""Gemini 每日上限觸發,降級到 188 CPU 備援 — 24h dedup每日重置"""
# 2026-04-26 critic-H1 hotfix by Claude Opus 4.7 — dedup key 加日期後綴