Skip to content
TxPipe Docs

Monitoring

Supernode includes a canonical monitoring stack as part of the control-plane.

Operators should treat Prometheus and Grafana as mandatory parts of the platform, not optional add-ons.

Monitoring layers

The intended monitoring workflow has three layers:

  1. the dashboard for fast workload health and logs
  2. Prometheus for metric collection and queryability
  3. Grafana for visualization and operator dashboards

Prometheus

Prometheus is the standard collection layer for Supernode.

Workloads that expose PodMonitor or related scrape targets are expected to integrate into this stack automatically once the control-plane is in place.

Use Prometheus when you need to:

  • confirm scrape targets
  • inspect raw time-series data
  • validate whether metrics are present before building dashboards
Terminal window
kubectl -n control-plane port-forward pod/prometheus-prometheus-0 9090:9090

Grafana

Grafana is the standard operator visualization layer.

Use Grafana when you want to:

  • inspect workload health over time
  • correlate sync, peer, and resource behavior
  • analyze producer-specific metrics
  • hand off dashboards to operators who do not need direct Prometheus queries
Terminal window
kubectl -n control-plane port-forward service/grafana 3000:3000

Dashboard fit

The Supernode dashboard should be your first stop.

Grafana should be the second stop when you need history, richer charts, or deeper workload context.

That is the intended operating rhythm for the platform.