8.2.1.5 Named Volume Backup

Named volume backup is the process of extracting a named volume's data into a portable archive, typically performed using a temporary container that mounts both the volume and a host location, providing a way to create durable backups independent of the volume's own ongoing existence.

The Standard Backup Pattern

A short-lived container mounts the volume to be backed up alongside a host directory, then archives the volume's contents into that host directory.

docker run --rm \
  -v app-data:/data:ro \
  -v $(pwd)/backups:/backup \
  alpine tar czf /backup/app-data-$(date +%Y%m%d).tar.gz -C /data .

This creates a timestamped archive of the volume's entire contents, written to a host directory where it can be retained, copied elsewhere, or otherwise managed independently of the volume itself.

Mounting the Volume Read-Only During Backup

Mounting the volume as read-only (:ro) during the backup process prevents the backup operation itself from accidentally modifying the data being backed up.

docker run --rm -v app-data:/data:ro -v $(pwd)/backups:/backup alpine tar czf /backup/backup.tar.gz -C /data .

Restoring From a Backup

The reverse operation extracts a previously created archive back into a volume, useful for restoring data after an incident or migrating data to a new environment.

docker run --rm \
  -v app-data-restored:/data \
  -v $(pwd)/backups:/backup \
  alpine tar xzf /backup/app-data-20260601.tar.gz -C /data

Automating Regular Backups

This same pattern can be scheduled to run periodically, ensuring regular, automated backups of important volume data without requiring manual intervention each time.

0 2 * * * docker run --rm -v app-data:/data:ro -v /backups:/backup alpine tar czf /backup/app-data-$(date +%Y%m%d).tar.gz -C /data .

Why Named Volume Backup Matters

Since a volume's persistence guards against container removal but not against host failure, disk corruption, or accidental volume deletion, establishing a reliable backup process is an essential complement to volume-based persistence for any data that genuinely cannot afford to be lost.