Skip to main content

Monitoring

Clutch Deploy includes Prometheus, Grafana, and Seq for metrics, dashboards, and structured logs.

Prometheus

  • URL: http://localhost:9090
  • Scrape targets: node1:3001, node2:3002, node3:3003
  • Config: config/monitoring/prometheus/prometheus.yml

Prometheus scrapes each node's /metrics endpoint on a fixed interval and stores time series for the Grafana dashboards.

Grafana

  • URL: http://localhost:3030 (port 3030 to avoid conflict with the API)
  • Login: admin / admin
  • Dashboards: Clutch Node dashboard (state, block index, latest blocks)
  • Config: config/monitoring/grafana/

Grafana dashboard

Dashboard panels typically cover:

MetricMeaning
Block indexHeight of the latest finalized block
Latest blocksRecent block hashes and authors
Node stateAura state (syncing, ready, etc.)
Connected peerslibp2p peer count
RPC rate / errorsJSON-RPC requests and failures per node

Metrics endpoints

TargetURL
Node 1http://localhost:3001/metrics
Node 2http://localhost:3002/metrics
Node 3http://localhost:3003/metrics

Seq (Logging)

Seq collects structured events (JSON) from nodes and the Hub API. Use it to trace transaction processing, RPC errors, and reconnection events. Protect it with SEQ_API_KEY if exposed beyond localhost.

Alert suggestions

ConditionSuggested action
Block index stops advancing on a validatorCheck Aura state and peer connectivity
RPC error rate > thresholdInspect Seq for malformed txs or node errors
Peer count drops to 0 on any nodeVerify libp2p ports and bootstrap peer
/health non-200 on Hub APIRestart clutch-hub-api, check node WS connection

Alert rules can be added to Prometheus via config/monitoring/prometheus/ and routed through Alertmanager if you wire one up.