monitoring.md

Monitoring (Prometheus + Grafana)

Start monitoring overlay

docker compose \
  -f docker-compose.yml \
  -f docker-compose.workers.yml \
  -f docker-compose.monitoring.yml \
  up -d

Use the workers overlay with monitoring when you want job metrics (jobs_run_total) and worker-level metrics from worker:9100.

Services:

• Prometheus: http://localhost:9090
• Grafana: http://localhost:3001

Metrics scraped

Prometheus scrapes:

• backend /metrics (backend:8000)
• worker /metrics (worker:9100) when worker metrics are enabled (provided by docker-compose.workers.yml)

Key metric families:

• http_requests_total
• http_request_duration_seconds
• matches_created_total
• matches_run_total
• tournaments_run_total
• identity_verifications_total
• recap_publishes_total
• jobs_run_total

Grafana dashboard

Provisioned dashboard: HowlHouse Overview

Panels:

• HTTP req/s
• HTTP p95 latency
• match/tournament run rates by status
• identity verification rate by outcome
• recap publish rate
• job run rate by type/status

Production note

Use production ingress controls for /metrics (basic auth/IP restrictions). See docs/deploy_production.md.