Self-hosted monitoring

active running since 2026-02-01
project · linux

Prometheus + Grafana across the whole home and lab infrastructure. The goal: see problems before they surface — and understand how the systems behave over time.

Architecture

Prometheus scrapes node_exporters and application metrics, Grafana visualises, Alertmanager routes notifications. Everything runs in containers on a single host with persistent volumes.

State and plans

  • covered: servers, network, Zigbee gateway, backups
  • in progress: alerting rules based on real incidents
  • planned: long-term storage (Thanos vs. VictoriaMetrics)

Timeline

02/2026 start — Prometheus + Grafana in containers
03/2026 alerting through Alertmanager → e-mail + ntfy
05/2026 dashboards for smart-home metrics (temperatures, consumption)

← back to explore