How-to

Monitor verify, prune, and GC for PBS.

Keep backups trustworthy by scheduling verify, prune, and GC with alerts and quick checks.

Steps

  1. Set verify cadence. Configure verify jobs per datastore (nightly/weekly depending on size). Stagger after backups.
  2. Schedule prune and GC. Run prune before GC, outside backup windows. Match retention policy.
  3. Enable notifications. Configure email/SMTP or webhook for job failures (verify, prune, GC) in PBS.
  4. Watch metrics. Track dedupe ratio, datastore utilization, and job durations. Add alerts for sudden duration spikes.
  5. Test restores monthly. Perform a restore drill from a recently verified snapshot and log time-to-restore.
  6. Document schedules. Keep a runbook with backup, verify, prune, GC, and sync windows to avoid overlap.

Prereqs

  • Defined retention policy
  • Email/webhook for alerts
  • Monitoring for PBS metrics
  • Runbook for schedules

Quick checks

  • Verify jobs succeed; failures alert you.
  • Prune runs before GC; no overlap with backups/sync.
  • Datastore free space stays within thresholds.
  • Monthly restore drill passes with documented timing.

If something fails

  • Increase verify concurrency gradually and re-run.
  • Adjust prune/GC timing to reduce contention.
  • Check chunk errors; re-verify and consider re-backup.
  • Expand datastore or tighten retention if space is low.

Related guides

Official docs

See Proxmox docs on verification and pruning: Prune/GC and Verify.

Hosted PBS at $7.95/TB.

No storage limits—$7.95/TB with compute and RAM included. We run the infrastructure; you keep control.

See pricing