:7778/metrics. This page lists the recommended dashboards, the alert thresholds operators run in production, and the integration points for status reporting.
Metric categories
| Category | Examples | Purpose |
|---|---|---|
| Consensus | sigil_consensus_round_total, sigil_consensus_rounds_missed_total | Track MACA participation. |
| Block apply | sigil_block_apply_seconds, sigil_state_root_compute_seconds | Track block-time budget. |
| Mempool | sigil_mempool_size, sigil_mempool_admitted_total | Track admission pressure. |
| P2P | sigil_p2p_peers_connected, sigil_p2p_bytes_in_total | Track network health. |
| Storage | sigil_akasha_compaction_seconds, sigil_akasha_disk_bytes | Track AkashaKV health. |
| RPC | sigil_rpc_requests_total, sigil_rpc_5xx_total | Track public surface load. |
Required alerts (validator)
| Alert | Threshold | Severity | Action |
|---|---|---|---|
| Validator inactive | Off the active set for > 1 epoch | Critical | Investigate immediately. |
| Round miss rate | > 5% over 24h | High | Investigate consensus link. |
| Equivocation detected | sigil_consensus_equivocation_total > 0 | Critical | Halt validator, audit keystore. |
| P99 block apply | > 4 s | High | Investigate CPU or I/O. |
| Disk usage | > 70% | High | Provision more storage or prune. |
| Peer count | < 8 | High | Investigate networking. |
| WAL lag | > 1 block | High | Investigate storage. |
| HSM signing latency | > 200 ms p99 | Medium | Investigate keystore. |
Required alerts (RPC node)
| Alert | Threshold | Severity | Action |
|---|---|---|---|
| Sync lag | > 10 blocks behind public RPC | High | Re-sync or restore from snapshot. |
| 5xx rate | > 1% over 5 minutes | High | Investigate. |
| 429 rate | > 10% over 5 minutes | Medium | Rate limit may need raising. |
| WebSocket queue overflow | sigil_ws_overflow_total > 0 | Medium | Investigate slow consumers. |
Grafana dashboards
The Sigil Foundation publishes reference dashboards:| Dashboard | Description |
|---|---|
sigil-validator-overview.json | One-pane validator overview. Round participation, block apply, peer health, slashing status. |
sigil-rpc-overview.json | RPC traffic, error rates, sync lag, subscription health. |
sigil-consensus.json | Per-round latency, proposer rotation, quorum margin. |
sigil-storage.json | AkashaKV compaction, WAL lag, disk usage. |
Status page integration
Production deployments should report to a status page. The Sigil Foundation uses Atlassian Statuspage; the validator cluster reports to status.sigil.ml. See the status page runbook for the integration recipe.Log aggregation
Validators emit structured JSON logs via thetracing crate. Recommended sinks: Loki, Elasticsearch, or Datadog. Critical fields:
level—INFO,WARN,ERROR.target— module path (sigil_consensus::round,sigil_rpc::handler, etc.).span— current span.block_height,round,validator_did— consensus context.