Attestor Observability Runbook

Audience: Attestor operators and SREs running the service in connected or air-gapped deployments.

Purpose: Describe the dashboards, metrics, logs, health endpoints, and alert thresholds used to monitor Stella Ops Attestor, including offline verification steps.

Dashboards (offline import)

Key metrics

MetricDescription
attestor.sign_latency_secondsSigning latency per request.
attestor.verify_total{result}Verification attempts, including failures by result.
attestor.submit_latency_secondsTransparency-log submission latency.
attestor.rekor_queue_depthPending/retrying Rekor submission jobs.
attestor.rekor_dead_letter_totalRekor submissions moved to dead letter.

Logs and traces

Health and diagnostics

Alert hints

ConditionSuggested threshold
Signing latency (p99)> 1s
Verification failuresSustained spike above baseline
Transparency log submission lag> 10s
Key rotation ageExceeds policy threshold
Queue backlogAbove configured threshold

Offline verification steps

  1. Import the Grafana JSON locally and point it at a Prometheus instance scraping the attestor target.
  2. Run a sample verification and confirm metrics and logs emit locally.
  3. Fetch /buildinfo.json and compare the commit, version, and image provenance against the offline bundle manifest.