Frigate Monitoring Dashboard

Frigate Monitoring Dashboard

This Grafana dashboard provides a complete observability view for Frigate using Prometheus as the data source.
It is designed for Frigate installations that expose metrics via the /api/metrics endpoint and covers runtime performance, object detection efficiency, and storage usage.

Features

• Python Runtime Monitoring

  • Garbage Collector (GC) objects collected/sec, collections per generation, uncollectable objects (leak detection), GC efficiency ratio.

• CPU & Memory Usage

  • Total system CPU/memory usage.
  • Per-process CPU/memory breakdown (ffmpeg, detect, capture, recording).

• Camera Performance

  • Camera input FPS vs processed FPS.
  • Detection FPS and skipped FPS (frame drops).

• Object Detection & Events

  • Detector inference speed (CPU/GPU).
  • Events per camera / per object label (if frigate_camera_events_total is exposed).

• Storage & Service Health

  • Storage usage (recordings, clips, cache, shm).
  • Service uptime.
  • Frigate version vs latest release (with update check).

Requirements

• Grafana ≥ 9.x
• Prometheus as datasource
• Frigate with metrics endpoint enabled (/api/metrics)

Prometheus Configuration

Add a scrape job for Frigate in your prometheus.yml:

YAML Copy

scrape_configs:
  - job_name: 'frigate'
    static_configs:
      - targets: ['192.168.1.50:5000']   # Replace with your Frigate host:port
    metrics_path: /api/metrics

Usage

1. Import this dashboard JSON into Grafana.
2. Select your Prometheus data source.
3. Adjust default time ranges if needed (recommended: 1h for CPU/FPS panels, 24h for storage panels).
4. (Optional) Configure alerts in Grafana for thresholds such as:
   • CPU usage > 85%
   • Memory usage > 90%
   • Storage usage > 80%
   • GC efficiency < 90%

Github Repo

Tags

frigate, cctv, surveillance, prometheus, grafana, monitoring, object-detection