feat(prometheus-direct-charting): slice 3 — remove grafana + config rewrite + changelog
Remove the entire Grafana surface: integrations/grafana.py, GrafanaWidgetSource (+ _fetch_chart, now redundant since prometheus chart exists), GrafanaLinkWidget, LinksTab, get_grafana_status endpoint, useGrafanaStatus hook, GrafanaStatus type, fetchGrafanaStatus client fn, registry/nav/tab entries (FE+BE). Rewrite config.yaml thin-dashboard rule to match reality (recharts is sanctioned for Prometheus-backed series). CHANGELOG migration note added. Backend: 293 pytest pass, ruff clean. Frontend: build+lint green (0 errors). SC-115/116 grep-clean (only prometheus_range.py migration comments + Dashboard.test shortcut fixture remain — both spec-allowed).
This commit is contained in:
Developer
@@ -1,7 +1,7 @@
|
||||
"""Base classes for service integrations.
|
||||
|
||||
A *service definition* is a closed, compile-time description of an external service
|
||||
the app can talk to (Grafana, Jellyfin, …). Each definition declares:
|
||||
the app can talk to (Jellyfin, Prometheus, …). Each definition declares:
|
||||
|
||||
* its non-secret ``config_schema`` (derived from a Pydantic model),
|
||||
* the secret fields it accepts (API keys / tokens),
|
||||
@@ -24,7 +24,7 @@ from pydantic import BaseModel, BeforeValidator, Field
|
||||
def _validate_service_base_url(value: Any) -> str:
|
||||
"""Require an absolute http(s) URL for service ``base_url`` fields.
|
||||
|
||||
Relative hosts (e.g. ``grafana.example.com``) break downstream HTTP clients
|
||||
Relative hosts (e.g. ``example.com``) break downstream HTTP clients
|
||||
because ``requests`` treats them as relative paths, so we fail fast with a
|
||||
clear error instead of letting the call silently malfunction.
|
||||
"""
|
||||
|
||||
@@ -1,73 +0,0 @@
|
||||
"""Grafana service definition."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from media_library_viewer_api.integrations.base import (
|
||||
SecretField,
|
||||
ServiceBaseUrl,
|
||||
ServiceConfigBase,
|
||||
ServiceDefinition,
|
||||
WidgetConfigBase,
|
||||
widget_kind,
|
||||
)
|
||||
|
||||
|
||||
class GrafanaConfig(ServiceConfigBase):
|
||||
"""Non-secret Grafana connection config."""
|
||||
|
||||
base_url: ServiceBaseUrl
|
||||
timeout_seconds: int = 5
|
||||
|
||||
|
||||
class GrafanaLinkWidgetConfig(WidgetConfigBase):
|
||||
"""Deep-link to a Grafana dashboard or panel."""
|
||||
|
||||
dashboard_uid: str
|
||||
panel_id: int | None = None
|
||||
|
||||
|
||||
class GrafanaChartWidgetConfig(WidgetConfigBase):
|
||||
"""Render a time-series chart from a Grafana datasource query."""
|
||||
|
||||
datasource_uid: str = "prometheus"
|
||||
query: str = ""
|
||||
from_ts: str = "now-1h"
|
||||
to_ts: str = "now"
|
||||
interval_ms: int = 30_000
|
||||
max_data_points: int = 100
|
||||
|
||||
|
||||
DEFINITION = ServiceDefinition(
|
||||
service_type="grafana",
|
||||
name="Grafana",
|
||||
description="Dashboards, metrics, and logs.",
|
||||
config_model=GrafanaConfig,
|
||||
secret_fields=[
|
||||
SecretField(key="api_key", label="API key", helper="Service account token (optional)"),
|
||||
],
|
||||
widget_kinds=[
|
||||
widget_kind(
|
||||
kind="link",
|
||||
name="Dashboard link",
|
||||
description="Deep-link to a Grafana dashboard or panel.",
|
||||
model_cls=GrafanaLinkWidgetConfig,
|
||||
default_config={"dashboard_uid": ""},
|
||||
refresh_interval_ms=0,
|
||||
),
|
||||
widget_kind(
|
||||
kind="chart",
|
||||
name="Chart",
|
||||
description="Live time-series chart from a Grafana datasource query.",
|
||||
model_cls=GrafanaChartWidgetConfig,
|
||||
default_config={
|
||||
"datasource_uid": "prometheus",
|
||||
"query": "",
|
||||
"from_ts": "now-1h",
|
||||
"to_ts": "now",
|
||||
"interval_ms": 30_000,
|
||||
"max_data_points": 100,
|
||||
},
|
||||
refresh_interval_ms=60_000,
|
||||
),
|
||||
],
|
||||
)
|
||||
@@ -10,14 +10,12 @@ from media_library_viewer_api.integrations.alertmanager import DEFINITION as ALE
|
||||
from media_library_viewer_api.integrations.authentik import DEFINITION as AUTHENTIK
|
||||
from media_library_viewer_api.integrations.backups import DEFINITION as BACKUPS
|
||||
from media_library_viewer_api.integrations.base import ServiceDefinition, WidgetKind
|
||||
from media_library_viewer_api.integrations.grafana import DEFINITION as GRAFANA
|
||||
from media_library_viewer_api.integrations.jellyfin import DEFINITION as JELLYFIN
|
||||
from media_library_viewer_api.integrations.nextcloud import DEFINITION as NEXTCLOUD
|
||||
from media_library_viewer_api.integrations.prometheus import DEFINITION as PROMETHEUS
|
||||
from media_library_viewer_api.integrations.ssh_tasks import DEFINITION as SSH_TASKS
|
||||
|
||||
SERVICE_DEFINITIONS: dict[str, ServiceDefinition] = {
|
||||
GRAFANA.service_type: GRAFANA,
|
||||
PROMETHEUS.service_type: PROMETHEUS,
|
||||
ALERTMANAGER.service_type: ALERTMANAGER,
|
||||
JELLYFIN.service_type: JELLYFIN,
|
||||
|
||||
@@ -2,7 +2,7 @@
|
||||
|
||||
Widgets are either:
|
||||
* **service-bound** — reference a ``service_id`` and a ``widget_kind`` declared
|
||||
by that service's definition (Grafana link, Prometheus metric, Jellyfin
|
||||
by that service's definition (Prometheus metric, Jellyfin
|
||||
activity, SSH task output); or
|
||||
* **built-in** — ``service_id`` is null and ``widget_kind`` is one of the
|
||||
service-less kinds (backups, static).
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
"""Monitoring router — observability service status.
|
||||
|
||||
Observability components (Alertmanager, Grafana, Prometheus) are resolved from
|
||||
Observability components (Alertmanager, Prometheus) are resolved from
|
||||
the service registry, not environment variables. The endpoints pick the first
|
||||
enabled instance of a type when no ``service_id`` is given, and return graceful
|
||||
"not configured" / "unreachable" payloads so the UI always renders a health card.
|
||||
@@ -172,29 +172,6 @@ def get_alertmanager_status(
|
||||
}
|
||||
|
||||
|
||||
@router.get("/grafana-status")
|
||||
def get_grafana_status(
|
||||
service_id: str | None = None,
|
||||
store: SettingsStore = Depends(get_settings_store),
|
||||
) -> dict[str, Any]:
|
||||
"""Probe a Grafana service instance's ``/api/health`` endpoint."""
|
||||
service = resolve_service_record(store, "grafana", service_id)
|
||||
if service is None:
|
||||
return _status_response(None, error="no_service_configured")
|
||||
try:
|
||||
response = requests.get(
|
||||
f"{_base_url(service)}/api/health",
|
||||
headers=_auth_headers(service),
|
||||
timeout=_timeout(service, 5),
|
||||
)
|
||||
response.raise_for_status()
|
||||
data = response.json()
|
||||
except Exception:
|
||||
logger.exception("Failed to fetch Grafana status")
|
||||
return _status_response(service, error="grafana_unreachable")
|
||||
return _status_response(service, version=data.get("version", ""))
|
||||
|
||||
|
||||
@router.get("/prometheus-status")
|
||||
def get_prometheus_status(
|
||||
service_id: str | None = None,
|
||||
|
||||
@@ -81,7 +81,7 @@ class SettingsStore:
|
||||
)
|
||||
conn.execute("CREATE INDEX IF NOT EXISTS idx_monitoring_machines_mode ON monitoring_machines(mode)")
|
||||
# The legacy SSH-scraping monitor (MonitoringPoller) was decommissioned;
|
||||
# metrics now live in Prometheus/node_exporter/Grafana. Drop the orphan
|
||||
# metrics now live in Prometheus/node_exporter. Drop the orphan
|
||||
# table on startup so existing databases get a clean slate.
|
||||
conn.execute("DROP TABLE IF EXISTS monitoring_machine_actions")
|
||||
conn.execute(
|
||||
|
||||
@@ -102,118 +102,6 @@ class StaticWidgetSource:
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class GrafanaWidgetSource:
|
||||
"""Build a Grafana deep-link or query datasource for a chart."""
|
||||
|
||||
async def fetch(self, service: ServiceRecord | None, widget_kind: str, config: dict[str, Any]) -> dict[str, Any]:
|
||||
try:
|
||||
if service is None:
|
||||
return {"error": "Grafana widget is missing its service"}
|
||||
base_url = str(service.config.get("base_url") or "").rstrip("/")
|
||||
api_key = str(service.secrets.get("api_key") or "")
|
||||
timeout = int(service.config.get("timeout_seconds") or 10)
|
||||
|
||||
if widget_kind == "chart":
|
||||
return await self._fetch_chart(base_url, api_key, timeout, config)
|
||||
|
||||
# Default: deep-link
|
||||
dashboard_uid = config.get("dashboard_uid")
|
||||
if not dashboard_uid:
|
||||
return {"error": "dashboard_uid is required"}
|
||||
url = f"{base_url}/d/{dashboard_uid}"
|
||||
panel_id = config.get("panel_id")
|
||||
if panel_id is not None:
|
||||
url = f"{url}?viewPanel={panel_id}"
|
||||
return {"url": url}
|
||||
except Exception as exc:
|
||||
logger.exception("grafana adapter failed")
|
||||
return {"error": f"Grafana link failed: {exc}"}
|
||||
|
||||
async def _fetch_chart(self, base_url: str, api_key: str, timeout: int, config: dict[str, Any]) -> dict[str, Any]:
|
||||
"""Query Grafana /api/ds/query and normalize to {series: [...]}."""
|
||||
if not api_key:
|
||||
return {"error": "Grafana api_key is required for chart queries"}
|
||||
query = config.get("query", "")
|
||||
if not query:
|
||||
return {"error": "query is required"}
|
||||
|
||||
datasource_uid = config.get("datasource_uid", "prometheus")
|
||||
body = {
|
||||
"queries": [
|
||||
{
|
||||
"datasource": {"uid": datasource_uid, "type": "prometheus"},
|
||||
"expr": query,
|
||||
"format": "time_series",
|
||||
"intervalMs": int(config.get("interval_ms", 30_000)),
|
||||
"maxDataPoints": int(config.get("max_data_points", 100)),
|
||||
"refId": "A",
|
||||
}
|
||||
],
|
||||
"from": config.get("from_ts", "now-1h"),
|
||||
"to": config.get("to_ts", "now"),
|
||||
}
|
||||
|
||||
def _do_post() -> dict[str, Any]:
|
||||
resp = requests.post(
|
||||
f"{base_url}/api/ds/query",
|
||||
json=body,
|
||||
headers={
|
||||
"Authorization": f"Bearer {api_key}",
|
||||
"Content-Type": "application/json",
|
||||
},
|
||||
timeout=timeout,
|
||||
)
|
||||
resp.raise_for_status()
|
||||
return resp.json()
|
||||
|
||||
try:
|
||||
raw = await asyncio.wait_for(asyncio.to_thread(_do_post), timeout=timeout)
|
||||
except asyncio.TimeoutError:
|
||||
return {"error": "Grafana query timed out"}
|
||||
except requests.RequestException as exc:
|
||||
return {"error": f"Grafana query failed: {exc}"}
|
||||
|
||||
# Normalize Grafana's /api/ds/query response into series.
|
||||
series: list[dict[str, Any]] = []
|
||||
results = raw.get("results", {})
|
||||
seen_labels: dict[str, int] = {}
|
||||
for ref_id, ref_data in results.items():
|
||||
for frame in ref_data.get("frames", []):
|
||||
values = frame.get("data", {}).get("values", [])
|
||||
if len(values) < 2:
|
||||
continue
|
||||
timestamps = values[0]
|
||||
vals = values[1]
|
||||
# Derive a meaningful series label from the frame metadata.
|
||||
# Prometheus frames carry metric labels in schema.fields[-1].labels.
|
||||
fields = frame.get("schema", {}).get("fields", [])
|
||||
value_field = fields[-1] if fields else {}
|
||||
# Prefer displayName (explicitly set in Grafana), then Prometheus
|
||||
# labels (e.g. {instance: "server:9100", mode: "iowait"}), then
|
||||
# the field name as a last resort.
|
||||
display_name = value_field.get("config", {}).get("displayName") or value_field.get("displayName")
|
||||
frame_labels = value_field.get("labels") or {}
|
||||
if display_name:
|
||||
label = str(display_name)
|
||||
elif frame_labels:
|
||||
# Build a readable label from the Prometheus labels, excluding
|
||||
# redundant ones like __name__.
|
||||
parts = [f"{k}={v}" for k, v in sorted(frame_labels.items()) if not k.startswith("__")]
|
||||
label = " ".join(parts) if parts else "value"
|
||||
else:
|
||||
label = value_field.get("name", "value")
|
||||
# Ensure unique labels when multiple series share the same name.
|
||||
if label in seen_labels:
|
||||
seen_labels[label] += 1
|
||||
label = f"{label} ({seen_labels[label]})"
|
||||
else:
|
||||
seen_labels[label] = 0
|
||||
points = [{"t": int(t), "v": float(v) if v is not None else None} for t, v in zip(timestamps, vals)]
|
||||
series.append({"label": label, "points": points})
|
||||
|
||||
return {"series": series}
|
||||
|
||||
|
||||
class PrometheusWidgetSource:
|
||||
"""Run PromQL queries against a Prometheus service (instant + range)."""
|
||||
|
||||
@@ -485,7 +373,6 @@ def _record_timeout(service: ServiceRecord | None, config: dict[str, Any], timeo
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
SERVICE_ADAPTERS: dict[str, WidgetSource] = {
|
||||
"grafana": GrafanaWidgetSource(),
|
||||
"prometheus": PrometheusWidgetSource(),
|
||||
"alertmanager": AlertmanagerWidgetSource(),
|
||||
"jellyfin": JellyfinWidgetSource(),
|
||||
|
||||
Reference in New Issue
Block a user