Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.superform.xyz/llms.txt

Use this file to discover all available pages before exploring further.

Use this page when you need validator observability and alerting guidance. For config fields, use Configuration Reference. For operator procedures, use Operations.

​
Health endpoints

Both ports are configured under [monitoring] in config.toml.
EndpointPortPurpose
GET /metricsmetrics_portPrometheus scrape target
GET /healthzhealth_portProcess + DB liveness
GET /readyzhealth_portSame checks today; reserved for future readiness gates
Quick check: 
curl -s localhost:8080/healthz
curl -s localhost:8080/readyz
curl -s localhost:9090/metrics | grep ^ocr2_ | head -20

​
Priority signals

If you only alert on a handful of metrics, start here:
  • ocr2_up — process liveness
  • ocr2_strategy_last_update_seconds — confirms PPS updates are still landing onchain
  • ocr2_strategy_update_stale — fastest signal that a strategy has stopped updating
  • ocr2_transmit_total labeled by status — shows whether transmit attempts are succeeding
  • ocr2_plugin_observations_total labeled by result — reveals RPC or observation failures
  • ocr2_config_version and ocr2_config_signers — confirm the node loaded the expected active network config

​
Metrics reference

​
Onchain transmission health

  • ocr2_strategy_last_update_seconds (labels: chain_id, strategy) — gauge with the Unix timestamp of the latest observed PPSValidated event
  • ocr2_strategy_update_stale (labels: chain_id, strategy) — gauge set to 1 when no PPSValidated event arrives within health_check_interval

​
Transmitter

  • ocr2_transmit_total (labels: chain_id, status) — counter of transmission attempts by outcome
  • ocr2_transmit_strategies_total (label: chain_id) — counter of strategies packed into submitted transactions
  • ocr2_transmit_pack_errors_total (label: chain_id) — counter of report packing failures
  • ocr2_transmit_duration_seconds (label: result) — histogram of end-to-end transmit latency

​
Plugin phases

  • ocr2_plugin_phase_duration_seconds (labels: phase, result) — histogram of phase latency
  • ocr2_plugin_phase_total (labels: phase, result) — counter of phase invocation outcomes
  • ocr2_plugin_observations_total (labels: chain_id, result) — counter of per-strategy observation outcomes
  • ocr2_plugin_insufficient_observations_total (label: chain_id) — counter of vaults dropped because quorum observations were not available

​
Alerting baseline

For production, alert on the smallest set of signals that prove the node can observe, participate, and transmit:
SignalSuggested action
ocr2_up missing or 0Page or notify the operator immediately
ocr2_strategy_update_stale == 1Treat as an incident; check RPC, config, and transmission logs
ocr2_transmit_total{status="failure"} increasingInvestigate transmitter key, RPC health, gas, and contract reverts
ocr2_plugin_observations_total{result="error"} increasingCheck per-chain RPC health and vault configuration
Unexpected ocr2_config_versionConfirm the node loaded the intended onchain config
Email, Slack, or SNS-style notifications are enough for most operators. Operators with SLA requirements should route these alerts into PagerDuty, Opsgenie, or an equivalent on-call system. After every config change, restart, or upgrade, verify three things in this order:
  1. /healthz is healthy
  2. ocr2_config_version and ocr2_config_signers match expectations
  3. ocr2_strategy_last_update_seconds continues advancing