feat(grafana-metric-gateway): slice 1 — backend gateway transport

Route all prometheus widget queries through Grafana /api/ds/query instead of
direct Prom HTTP. PrometheusConfig: drop base_url, add grafana_url +
datasource_uid; secret grafana_api_key (required). PrometheusWidgetSource →
MetricSource with _gateway_query POST method. normalize_grafana_frames
recovered from 65bae95 + shared _dedup_label helper. Gateway-path status
check. Startup old-config validation. CHANGELOG migration note. All adapter
tests rewritten for POST /api/ds/query + Grafana frames mock. Backend: 331
pytest pass, ruff clean. Frontend: build green (unchanged in S1).
This commit is contained in:
Developer
2026-07-09 21:24:44 +00:00
parent 798196ffc7
commit df80c68f89
10 changed files with 573 additions and 250 deletions
@@ -177,23 +177,52 @@ def get_prometheus_status(
service_id: str | None = None,
store: SettingsStore = Depends(get_settings_store),
) -> dict[str, Any]:
"""Probe a Prometheus service instance's health and build info."""
"""Probe a Prometheus service's health via the Grafana gateway path (GM-110).
Issues a trivial ``up`` query through Grafana ``/api/ds/query``. Success
validates the full path: Grafana is reachable, the API key works, and the
Prometheus datasource responds.
"""
service = resolve_service_record(store, "prometheus", service_id)
if service is None:
return _status_response(None, error="no_service_configured")
base = _base_url(service)
timeout = _timeout(service, 10)
headers = _auth_headers(service)
grafana_url = str(service.config.get("grafana_url") or "").rstrip("/")
api_key = str(service.secrets.get("grafana_api_key") or "")
datasource_uid = str(service.config.get("datasource_uid") or "prometheus")
timeout = int(service.config.get("timeout_seconds") or 10)
if not grafana_url or not api_key:
return _status_response(service, error="gateway_not_configured")
body = {
"queries": [
{
"datasource": {"uid": datasource_uid, "type": "prometheus"},
"expr": "up",
"format": "time_series",
"intervalMs": 15_000,
"maxDataPoints": 1,
"refId": "A",
}
],
"from": "now-1m",
"to": "now",
}
try:
health = requests.get(f"{base}/-/healthy", headers=headers, timeout=timeout)
health.raise_for_status()
build_info = requests.get(f"{base}/api/v1/status/buildinfo", headers=headers, timeout=timeout)
build_info.raise_for_status()
version = build_info.json().get("data", {}).get("version", "")
except Exception:
logger.exception("Failed to fetch Prometheus status")
resp = requests.post(
f"{grafana_url}/api/ds/query",
json=body,
headers={"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"},
timeout=timeout,
)
resp.raise_for_status()
except requests.HTTPError as exc:
status_code = exc.response.status_code if exc.response else 0
if status_code in (401, 403):
return _status_response(service, error="auth_failed")
return _status_response(service, error="gateway_error")
except requests.RequestException:
logger.exception("Failed to fetch Prometheus status via gateway")
return _status_response(service, error="prometheus_unreachable")
return _status_response(service, version=version)
return _status_response(service, version="ok")
@router.post("/alertmanager-webhook")