In the n8n-shortlink-network network, we have these services running:
- http://n8n-shortlink:3001 → Shortlink API exposes Prometheus metrics, some derived from expvar.
- http://prometheus:9090 → Prometheus to scrape and aggregate metrics.
- http://node_exporter:9100 → Node Exporter to expose host metrics.
- http://cadvisor:8080 → cAdvisor to expose container metrics.
- http://loki:3100 → Loki to aggregate logs from Promtail.
- Promtail (no port exposed) to push logs to Loki.
- http://grafana:3000 → Grafana to visualize metrics and logs.
Prometheus aggregates metrics from itself, the API, node exporter, cAdvisor. Loki aggregates logs from Promtail, which pushes them from the API and from the backup cronjob. Grafana visualizes metrics and logs at grafana.n8n.to. Outside the network, e.g. to the reverse proxy, these services are available as localhost:{port}.
- Set up Grafana dashboards
- Set up UptimeRobot or Checkly
- Set up https://healthchecks.io/ for backup cronjob
- Monitor sqlite DB (how?)
- Makefile commands for server logs
- Makefile commands for backups
- Makefile command for Sentry