Skip to main content

Backup & Restore overview

Backup on FPT MongoDB Enterprise is built from four pieces. Unlike some engines, the first two are already switched on for you — but that still does not mean your data is protected.

PieceWhat it doesOn its own
Backup serviceReserves storage capacity and holds the policy every job usesBacks up nothing
Backup jobThe schedule that actually produces restore pointsThis is what creates recoverable data
PITRRetains archived logs so you can recover to an exact momentNeeds a full backup, and takes effect from the next one onward
RestoreCreates a new cluster from a restore point or a moment in timeNever overwrites the original
warning

Backup and PITR are enabled by default when a MongoDB Enterprise cluster is created, but no data is protected until a backup job exists and has completed at least one run. Check the Backup tab of a new cluster before assuming it is covered.

What is on by default

When you provision a MongoDB Enterprise cluster, the backup and PITR services are enabled as part of creation. This differs from other FPT Database Engine engines, where both need enabling explicitly.

What is not automatic is the backup job — the schedule that produces restore points. Confirm one exists and has run.

Where to go next

You want toGo to
Check or change the backup schedule and retentionChange backup schedule
Recover the state captured by a backupRestore from backup
Recover the state at an exact momentRecover to point in time

Recovery always creates a new cluster

Both recovery paths build a new cluster rather than repairing the existing one. Two consequences worth planning for:

  • You need quota for a second cluster at the very moment you are recovering, which is the worst possible time to discover you are at your limit.
  • Afterwards, repoint your application. Recovery never moves the endpoint for you.

What it costs

Backup storage is billed separately from the cluster. PITR adds to that, because archived logs are retained alongside the backups. Both are worth estimating up front, since capacity applies across every backup job on the cluster rather than per job.

Next steps