Monitoring¶

Prometheus Metrics¶

Hafiz exposes metrics at /metrics on the main S3 port (9000 by default).

Access Metrics¶

curl http://localhost:9000/metrics

Key Metrics¶

Metric	Type	Description
`hafiz_requests_total`	Counter	Total requests
`hafiz_request_duration_seconds`	Histogram	Request latency
`hafiz_objects_total`	Gauge	Object count
`hafiz_storage_bytes`	Gauge	Storage used
`hafiz_active_connections`	Gauge	Active connections

Prometheus Config¶

scrape_configs:
  - job_name: 'hafiz'
    static_configs:
      - targets: ['hafiz:9000']
    metrics_path: '/metrics'

Grafana Dashboard¶

Import the dashboard from:

deploy/grafana/dashboards/hafiz.json

Alerts¶

Example PrometheusRule:

groups:
  - name: hafiz
    rules:
      - alert: HafizHighErrorRate
        expr: rate(hafiz_requests_total{status="error"}[5m]) > 0.1
        for: 5m
        labels:
          severity: warning

Health Checks¶

# Health
curl http://localhost:9000/health

# Readiness
curl http://localhost:9000/ready

# Liveness
curl http://localhost:9000/live