Monitoring

We want to remotely monitor the health, activity, and related metrics of

• back-end application code
• web server, e.g. Nginx (production) or Uvicorn (development)
• database
• front-end

For the first three items, the initial implementation will use Prometheus for metrics and Grafana for visual display of the data.

Monitoring Components and URLs

Component	Monitor	URL
FastAPI WS	prometheus-fastapi-instrumentor + custom code in main.py	http://localhost:8080/metrics
Nginx	/stub_status, accessible only inside Docker network	http://nginx:8080/stub_status
Nginx Exportor	nginx-exporter running in a separate container exposed port 9113	http://localhost:9113/metrics
Prometheus Server & UI	Collects ("scrapes") metrics	http://localhost:9090
Grafana UI	http://localhost:3000

Note that Docker resolves container names like nginx to IP addresses. Use docker network ls to discover network names, docker network inspect homelog_default to discover container IP addresses.

As of this writing, the relevant files and directories are:

project/
├── docker-compose.yml    Configure containers, exposed ports, and volumes
├── backend/
│   ├── Dockerfile        FastAPI web service + Uvicorn
│   └── main.py           Configure and expose metrics for Prometheus
├── nginx/                Nginx proxies the backend web service
│   └── nginx.conf        Exposes metrics at path `/stub_status`
└── monitor/
    └── prometheus.yml