Prometheus
Metrics storage and querying for BlumeOps infrastructure.
Quick Reference
| Property | Value |
|---|---|
| URL | https://prometheus.ops.eblu.me |
| Namespace | monitoring |
| Image | registry.ops.eblu.me/blumeops/prometheus (see argocd/manifests/prometheus-ringtail/kustomization.yaml for current tag) |
| Storage | 20Gi PVC |
| Manifests | argocd/manifests/prometheus-ringtail/ |
| ArgoCD app | prometheus-ringtail |
Data Sources
Remote Write (from Alloy)
- Indri system metrics via Alloy remote_write
- Ringtail Kubernetes metrics (kube-state-metrics, node, kubelet) via the alloy-ringtail DaemonSet
- Textfile metrics: borgmatic, zot, jellyfin, forgejo
- flyio-proxy nginx metrics (
flyio_nginx_*) via Alloy embedded in the Fly.io container
Scrape Targets
| Job | Target |
|---|---|
node-exporter-sifaka | nas.ops.eblu.me:9100 — sifaka NAS (via Caddy L4 on indri) |
smartctl-sifaka | nas.ops.eblu.me:9633 — sifaka NAS (via Caddy L4 on indri) |
cnpg-postgres | blumeops-pg-metrics.databases.svc:9187, immich-pg-metrics.databases.svc:9187 — postgresql |
prometheus | localhost:9090 — self-monitoring |
loki | loki.monitoring.svc:3100 |
tempo | tempo.monitoring.svc:3200 |
argocd | argocd-metrics.argocd.svc:8082 |
frigate | nvr.ops.eblu.me/api/metrics |
birdnet-go | birdnet-go.birdnet-go.svc:8090 |
kube-state-metrics is deliberately not scraped here: alloy-ringtail scrapes and remote-writes it instead.