v0.2.0

Adds an optional dead-man's-switch heartbeat.

Set HEARTBEAT_URL and PVEWatch pings it after every successful poll cycle. Point it at healthchecks.io or an Uptime Kuma push monitor and you get alerted when PVEWatch goes silent, whether the container died or Proxmox became unreachable. Non-2xx ping responses are logged so a mistyped URL is visible immediately.

Also new since v0.1.0: WEB_UI_USERNAME/WEB_UI_PASSWORD are documented in .env.example, and the README gained development setup notes.

Images: ghcr.io/christianklass/pvewatch:0.2.0 (amd64, arm64)

v0.1.0

First public release.

PVEWatch watches your Proxmox VE cluster and alerts when backups fail, VMs go down, or storage crosses a threshold. Single container, no Prometheus or Grafana required.

• Per-VM backup history heatmap (3 to 30 days) in a read-only web dashboard
• Batch ("backup all") jobs parsed per VM from the task log
• Email and Discord alerts with the failing task's log tail
• Weekly digest, storage threshold alerts, Prometheus /metrics endpoint
• Optional HTTP Basic Auth for the dashboard and API
• Multi-node clusters auto-discovered; SQLite by default, Postgres optional

Images: ghcr.io/christianklass/pvewatch:0.1.0 (amd64, arm64)