Skip to content

Monitoring & Metrics

kiok surfaces the state of the cluster and of every workflow run through the admin UI.

  • Dashboard — aggregate run metrics: counts by state (success / failed / running), throughput over time, and recent activity.
  • Topology — the live cluster snapshot: the leader Master, every registered Master and Worker, per-Worker task slots, and readiness.
  • Run views — per run, a graph view (each task coloured by state), a job list (per-task worker, timing, exit code), and a timeline.
  • Health & readiness endpoints — /healthz (liveness) and /readyz (returns 200 only when fully ready) for container probes and external monitoring.
  • Logs — each role writes a rolling log file under logs/; per-task and per-run logs are streamed live and retained — see Job Logs.