Stronghold backups are based on snapshots of integrated Raft storage. A snapshot lets you save cluster state and later use it to restore data after a failure or during planned maintenance operations.

Stronghold supports two main scenarios:

  • Manual snapshots: An administrator explicitly saves a snapshot file and restores it manually if needed.
  • Automated snapshots: Stronghold saves backups on a schedule to the local disk or S3-compatible storage.

Using snapshots

Snapshots are useful for:

  • Protection against data corruption or operator mistakes.
  • Preparation for upgrades and other risky operations.
  • Disaster recovery.
  • Storing backups outside the cluster or hosting site.

Consider the following when working with snapshots:

  • Snapshots apply to Stronghold integrated Raft storage.
  • Snapshots work only for clusters that use integrated Raft storage as the primary storage backend.
  • If Stronghold uses etcd, PostgreSQL, or another external system as a backend, you have to use the corresponding backup mechanisms of those systems, rather than the built-in Stronghold commands for Raft snapshots (d8 stronghold operator raft snapshot).
  • Data remains encrypted in the backup.
  • To regain access after restore, you need the correct unseal or recovery keys according to your cluster configuration.
  • Automated snapshots are an administrative feature and require a deliberate retention policy.

Recommendations

Follow these recommendations when using Stronghold backups:

  • Keep backups outside the same environment they are intended to protect.
  • Periodically verify snapshot creation with d8 stronghold operator raft snapshot inspect and the backup restore procedure in a test environment.
  • For production environments, prefer external object storage over local disk whenever possible.

Additional resources