Create a volume backup
Volume backups are a deeper recovery mechanism than snapshots — they're a full copy of the volume that can be restored independently and stored on a separate backend.
The Portal doesn't surface a volume-backup UI today; create and manage backups from the OpenStack CLI.
Backup vs snapshot
| Snapshot | Backup |
|---|---|
| Cheap, fast, near-instant. | Slower, copies all blocks. |
| Lives on the same storage backend as the volume. | Can target a separate backup backend. |
| Lost if the source backend fails catastrophically. | Survives source-backend loss. |
| Pointed in time, restored via "create volume from snapshot". | Full restore, replaces or branches a volume. |
| Managed from the Portal. | Managed via the CLI. |
A typical pattern: take frequent snapshots for fast rollback, take less-frequent backups for disaster recovery.
Prerequisites
- An existing volume in any state.
- The OpenStack CLI
installed and pointed at a
clouds.yamlwith an application credential for the project that owns the volume. - Operator-side volume-backup configuration. If your deployment
hasn't been configured with a backup backend,
openstack volume backup createwill return an error — contact support to confirm backup is available before you build automation around it.
Steps
1. Look up the volume ID
Note the ID column for the volume you want to back up.
2. Create the backup
For volumes that are currently attached (in-use), add --force:
The command returns the backup's ID and name:
+-------+--------------------------------------+
| Field | Value |
+-------+--------------------------------------+
| id | 2ac93a23-667d-4b4f-a7ba-d1778c3c21d0 |
| name | None |
+-------+--------------------------------------+
Add --name <label> and --description <text> to make backups
easier to identify later:
openstack --os-cloud breqwatr volume backup create <volume-id> \
--name pg-data-2026-04-19 \
--description "Before Postgres 16 upgrade" \
--force
3. Check status and list
openstack --os-cloud breqwatr volume backup list
openstack --os-cloud breqwatr volume backup show <backup-id>
Backups move through creating → available. Backup creation can take minutes for large volumes — the duration scales with volume size, not with how much data is on it.
4. Restore a backup
Restore creates a new volume (you can also restore over an existing volume, but creating a new one is safer):
Then attach the restored volume to an instance and verify the data.
5. Delete an old backup
Deleting is irreversible — verify you don't need the backup before running this.
Verification
openstack volume backup show <backup-id>returns the backup in available state.- Restoring the backup into a new volume produces data that matches the source (compare a directory listing or file checksum).
Next steps
- Take a snapshot for the fast-recovery side of the same protection story.
- Resize a volume — backups are a sensible pre-resize safety net for irreplaceable data.
- Use the OpenStack CLI if this is your first time running CLI commands against the cloud.