grafana
Catalog date: 2026-06-24.
The platform's primary dashboard surface. Bundled with the kube-prometheus-stack and pre-loaded with datasources for Prometheus, VictoriaMetrics, Loki (logs), and Jaeger (traces).
Identity
| Field | Value |
|---|---|
| Service id | grafana |
| Role | observability |
| Image | grafana/grafana (managed by kube-prometheus-stack) |
| Port | 3000 |
| Health | /api/health |
Deployment surfaces
| Surface | Where |
|---|---|
| Kustomize | folded into observability/kube-prometheus-stack/ (Helm chart bundles Grafana) |
| Standalone | (none — Grafana is always shipped with the stack) |
Datasources
Per grafana-datasources.yaml
(sidecar-discovered via the grafana_datasource label):
| Datasource | Backend |
|---|---|
Prometheus | prometheus-operated:9090 (in-cluster, default) |
Loki | loki:3100 (in-cluster) |
Jaeger | jaeger-query:16686 (in-cluster) |
OpenSearch Logs | opensearch:9200 |
QuestDB | questdb:8812 (Postgres-protocol datasource) |
Phoenix UI | phoenix:6006 — a prometheus-type datasource against Phoenix's own endpoint, not a Postgres backend |
No VictoriaMetrics datasource block was found in this file — flagging
the earlier "VictoriaMetrics" row as unverified/likely stale.
Dashboards
Provisioned via ConfigMaps carrying the grafana_dashboard label — the
Helm chart's dashboard sidecar (dashboards.enabled: true,
searchNamespace: ALL) discovers them cluster-wide. No
observability/kube-prometheus-stack/dashboards/ directory or any
grafana_dashboard-labeled ConfigMap currently exists in
alphaswarm_platform, so the previously listed default dashboard set
(cluster health, AlphaSwarm runtime, Linkerd proxy metrics, per-cell
tenant overlays) could not be verified — flagging as unverified rather
than asserting specific shipped dashboards.
Custom dashboards land via PR — never via the Grafana UI alone (UI edits are wiped on the next reconciliation).
Operations
- Auth: per
helm-values.yaml, Grafana'sauth.generic_oauthblock is currently wired to Auth0 (alphaswarm-fund.us.auth0.com), not an Entra tenant, withauto_assign_org_role: Editorfor every authenticatedalpha-swarm.aiuser — there is noalphaswarm-staff/alphaswarm-operatorsrole-mapping config in that file. Correcting the earlier claim of an Entra-backed admin/editor/viewer split, which does not match the current chart config. - Persistence: Grafana DB is SQLite by default (folded into the Helm chart); production cells point at a per-cell Postgres schema.
See also
prometheus.md,victoriametrics.md,loki.md,jaeger.md— backing datasources.observability-stack.md— stack composition.