How-to

Plan backup windows and concurrency.

Lay out backup, verify, prune, GC, and sync schedules with safe concurrency so jobs finish predictably.

Steps

  1. Map job durations. Measure typical backup, verify, prune, and GC times per datastore.
  2. Set a backup window. Choose hours with lowest production load; schedule backups first.
  3. Stagger verify. Run verify after backups, avoiding overlap with prune/GC and heavy IO.
  4. Place prune then GC. Run prune before GC, outside backups/verify; stagger per datastore.
  5. Limit concurrency. Cap simultaneous backups per datastore to match storage/network capacity.
  6. Review monthly. Adjust schedules as data grows or windows change.

Prereqs

  • Observed job durations
  • Target backup window
  • Datastore performance profile
  • Retention policy defined

Quick checks

  • No overlap between backups, verify, prune, and GC.
  • Job durations fit inside planned windows.
  • Concurrency limits prevent I/O spikes.
  • Schedules documented and monitored.

If something fails

  • Reduce concurrent jobs or stagger per datastore.
  • Move verify to lighter hours; reduce concurrency.
  • Adjust prune/GC timing or retention to free space.
  • Monitor storage/network for bottlenecks and resize.

Related guides

Official docs

See Proxmox VE backup scheduling guidance: PVE backup settings.

Hosted PBS at $7.95/TB.

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

See pricing