Skip to content

Restoring your data

Rested restores a snapshot by queuing a command that your agent runs on the server. You can drive it entirely from the dashboard, or run the equivalent command yourself.

From the dashboard

  • Open the repository and select the Snapshots tab. If the list is empty, press Refresh while the agent is online.
  • Pick a snapshot and choose Restore.
  • Set a target path at or below the administrator-owned /var/restore directory on the server. The default is a fresh directory outside your original data, so a restore does not overwrite live files by default. Restores to protected system locations such as /etc, /usr, /root, and /var/lib are rejected, including when a target reaches them through a symlink.
  • Choose a restore type. Restore files / artifacts writes the backed-up files (and any database dump) to the target path and is non-destructive. For database recipes, Auto-restore into the database additionally imports the dump back into the live database — this overwrites it, so it asks you to confirm.
  • Queue the restore. The agent picks it up on its next check-in; the result and logs appear in the repository’s history.

Running it yourself

Every restore is a standard restic operation, so you can run it directly on the server instead. The dashboard restore dialog shows the exact command under “Prefer to run it yourself?”. It looks like:

sudo rested restore --source <source-id> --snapshot <snapshot-id> --target /var/restore/rested/manual

Add --auto-database to import a database dump back into its configured database after the files are restored.

If you have neither the dashboard nor the agent available, you can restore with plain restic and your recovery key — see Recovery & your encryption key.