feat(monitoring): 監控工具區塊 — Grafana/Prometheus/SigNoz/Gitea 狀態
- 新增 GET /api/v1/monitoring/status，asyncio.gather 並行探測四工具 - 前端 MonitoringTools 元件，60s 輪詢顯示狀態/版本/統計 - 新增 monitoringTools i18n key Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
OG T
174
apps/api/src/api/v1/monitoring.py
Normal file
174
apps/api/src/api/v1/monitoring.py
Normal file
@@ -0,0 +1,174 @@
|
||||
"""
|
||||
Monitoring Status API
|
||||
=====================
|
||||
探測各監控工具狀態：Grafana / Prometheus / SigNoz / Gitea
|
||||
|
||||
所有探測從後端發出，不暴露內網 IP 給前端。
|
||||
|
||||
建立時間: 2026-04-03 (台北時區)
|
||||
建立者: Claude Code
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
from datetime import UTC, datetime
|
||||
|
||||
import httpx
|
||||
from fastapi import APIRouter
|
||||
|
||||
from src.core.logging import get_logger
|
||||
|
||||
logger = get_logger(__name__)
|
||||
|
||||
router = APIRouter(prefix="/monitoring", tags=["Monitoring"])
|
||||
|
||||
# =============================================================================
|
||||
# Internal service endpoints (backend-only)
|
||||
# =============================================================================
|
||||
|
||||
SERVICES = {
|
||||
"grafana": {
|
||||
"base": "http://192.168.0.188:3000",
|
||||
"health": "/api/health",
|
||||
"build": "/api/frontend/settings",
|
||||
},
|
||||
"prometheus": {
|
||||
"base": "http://192.168.0.110:9090",
|
||||
"health": "/-/healthy",
|
||||
"build": "/api/v1/status/buildinfo",
|
||||
"rules": "/api/v1/rules",
|
||||
},
|
||||
"signoz": {
|
||||
"base": "http://192.168.0.188:3301",
|
||||
"health": "/api/v1/health",
|
||||
},
|
||||
"gitea": {
|
||||
"base": "http://192.168.0.110:3001",
|
||||
"health": "/-/readiness",
|
||||
},
|
||||
}
|
||||
|
||||
TIMEOUT = 3.0
|
||||
|
||||
|
||||
async def _probe_grafana(client: httpx.AsyncClient) -> dict:
|
||||
base = SERVICES["grafana"]["base"]
|
||||
try:
|
||||
r = await client.get(f"{base}/api/health", timeout=TIMEOUT)
|
||||
if r.status_code == 200:
|
||||
data = r.json()
|
||||
version = data.get("version", "—")
|
||||
# Try to get dashboard count
|
||||
dash_r = await client.get(f"{base}/api/search?type=dash-db", timeout=TIMEOUT)
|
||||
dash_count = len(dash_r.json()) if dash_r.status_code == 200 else None
|
||||
return {
|
||||
"name": "Grafana",
|
||||
"status": "up",
|
||||
"version": version,
|
||||
"stats": f"面板 {dash_count} 個" if dash_count is not None else None,
|
||||
"description": "監控面板 · 指標視覺化",
|
||||
}
|
||||
except Exception as e:
|
||||
logger.warning("grafana_probe_failed", error=str(e))
|
||||
return {"name": "Grafana", "status": "down", "version": None, "stats": None, "description": "監控面板 · 指標視覺化"}
|
||||
|
||||
|
||||
async def _probe_prometheus(client: httpx.AsyncClient) -> dict:
|
||||
base = SERVICES["prometheus"]["base"]
|
||||
try:
|
||||
health_r = await client.get(f"{base}/-/healthy", timeout=TIMEOUT)
|
||||
if health_r.status_code == 200:
|
||||
# Get build info
|
||||
build_r = await client.get(f"{base}/api/v1/status/buildinfo", timeout=TIMEOUT)
|
||||
version = None
|
||||
if build_r.status_code == 200:
|
||||
version = build_r.json().get("data", {}).get("version")
|
||||
# Get rules count
|
||||
rules_r = await client.get(f"{base}/api/v1/rules", timeout=TIMEOUT)
|
||||
rules_count = None
|
||||
firing_count = 0
|
||||
if rules_r.status_code == 200:
|
||||
groups = rules_r.json().get("data", {}).get("groups", [])
|
||||
rules_count = sum(len(g.get("rules", [])) for g in groups)
|
||||
firing_count = sum(
|
||||
1 for g in groups for r in g.get("rules", [])
|
||||
if r.get("state") == "firing"
|
||||
)
|
||||
stats_parts = []
|
||||
if rules_count is not None:
|
||||
stats_parts.append(f"規則 {rules_count} 條")
|
||||
if firing_count > 0:
|
||||
stats_parts.append(f"{firing_count} 觸發")
|
||||
return {
|
||||
"name": "Prometheus",
|
||||
"status": "up",
|
||||
"version": version,
|
||||
"stats": " · ".join(stats_parts) if stats_parts else None,
|
||||
"description": "時序資料庫 · 告警規則",
|
||||
"firing_count": firing_count,
|
||||
}
|
||||
except Exception as e:
|
||||
logger.warning("prometheus_probe_failed", error=str(e))
|
||||
return {"name": "Prometheus", "status": "down", "version": None, "stats": None, "description": "時序資料庫 · 告警規則", "firing_count": 0}
|
||||
|
||||
|
||||
async def _probe_signoz(client: httpx.AsyncClient) -> dict:
|
||||
base = SERVICES["signoz"]["base"]
|
||||
try:
|
||||
r = await client.get(f"{base}/api/v1/health", timeout=TIMEOUT)
|
||||
if r.status_code == 200:
|
||||
return {"name": "SigNoz", "status": "up", "version": None, "stats": "APM · 追蹤 · 日誌", "description": "可觀測性平台"}
|
||||
except Exception as e:
|
||||
logger.warning("signoz_probe_failed", error=str(e))
|
||||
# Fallback: try root
|
||||
try:
|
||||
r2 = await client.get(f"{base}/", timeout=TIMEOUT)
|
||||
if r2.status_code in (200, 301, 302):
|
||||
return {"name": "SigNoz", "status": "up", "version": None, "stats": "APM · 追蹤 · 日誌", "description": "可觀測性平台"}
|
||||
except Exception:
|
||||
pass
|
||||
return {"name": "SigNoz", "status": "down", "version": None, "stats": None, "description": "可觀測性平台"}
|
||||
|
||||
|
||||
async def _probe_gitea(client: httpx.AsyncClient) -> dict:
|
||||
base = SERVICES["gitea"]["base"]
|
||||
try:
|
||||
r = await client.get(f"{base}/-/readiness", timeout=TIMEOUT)
|
||||
if r.status_code == 200:
|
||||
# Get version from API
|
||||
ver_r = await client.get(f"{base}/api/v1/version", timeout=TIMEOUT)
|
||||
version = None
|
||||
if ver_r.status_code == 200:
|
||||
version = ver_r.json().get("version")
|
||||
return {"name": "Gitea", "status": "up", "version": version, "stats": "CI/CD · Git 倉庫", "description": "代碼倉庫 · Pipeline"}
|
||||
except Exception as e:
|
||||
logger.warning("gitea_probe_failed", error=str(e))
|
||||
return {"name": "Gitea", "status": "down", "version": None, "stats": None, "description": "代碼倉庫 · Pipeline"}
|
||||
|
||||
|
||||
@router.get("/status")
|
||||
async def get_monitoring_status() -> dict:
|
||||
"""
|
||||
並行探測所有監控工具狀態
|
||||
|
||||
Returns:
|
||||
dict with tools list, each containing name/status/version/stats/description
|
||||
"""
|
||||
async with httpx.AsyncClient(follow_redirects=True) as client:
|
||||
results = await asyncio.gather(
|
||||
_probe_grafana(client),
|
||||
_probe_prometheus(client),
|
||||
_probe_signoz(client),
|
||||
_probe_gitea(client),
|
||||
return_exceptions=True,
|
||||
)
|
||||
|
||||
tools = []
|
||||
for r in results:
|
||||
if isinstance(r, Exception):
|
||||
continue
|
||||
tools.append({**r, "checked_at": datetime.now(UTC).isoformat()})
|
||||
|
||||
return {
|
||||
"tools": tools,
|
||||
"checked_at": datetime.now(UTC).isoformat(),
|
||||
}
|
||||
@@ -57,6 +57,7 @@ from src.api.v1 import (
|
||||
from src.api.v1 import (
|
||||
signoz_webhook as signoz_webhook_v1, # Phase 21: SignOz → Telegram (ADR-037)
|
||||
)
|
||||
from src.api.v1 import monitoring as monitoring_v1 # 2026-04-03: 監控工具狀態
|
||||
from src.api.v1 import stats as stats_v1 # Phase 6.5: Statistics Analytics
|
||||
from src.api.v1 import telegram as telegram_v1 # Phase 5.4: Telegram Gateway
|
||||
from src.api.v1 import terminal as terminal_v1 # Phase 19.1: Omni-Terminal SSE
|
||||
@@ -402,6 +403,9 @@ app.include_router(
|
||||
app.include_router(
|
||||
stats_v1.router, prefix="/api/v1", tags=["Statistics"]
|
||||
) # Phase 6.5: Statistics Analytics
|
||||
app.include_router(
|
||||
monitoring_v1.router, prefix="/api/v1", tags=["Monitoring"]
|
||||
) # 2026-04-03: 監控工具狀態
|
||||
app.include_router(
|
||||
github_webhook_v1.router, prefix="/api/v1", tags=["GitHub Webhook"]
|
||||
) # Phase 13.1: GitHub → OpenClaw
|
||||
|
||||
Reference in New Issue
Block a user