Backup & restore

Use the backup commands with the configuration and data directory of the instance to protect. The CLI reads local configured storage; running it on an unrelated machine does not back up your remote server.

Create a compressed backup

anubis backup create \
  --include-history \
  --output ./anubis-backup.json.gz

Compression is enabled by default. --include-history includes judgment history. Use --no-compress for an uncompressed output.

List and inspect

anubis backup list
anubis backup info ./anubis-backup.json.gz

Inspect the backup before a restore. Store copies separately from the machine and volume they protect.

Restore during maintenance

Plan a maintenance window and preserve the current data before replacing it:

anubis restore ./anubis-backup.json.gz --force

--force permits replacing existing data. Confirm the file is the backup you intend to restore and use the matching instance configuration.

Verify recovery

After restarting:

anubis health
anubis status
anubis judge --all

Check the dashboard, targets and alert channels. Rehearse recovery rather than treating a backup file's existence as proof of a successful restore.

See the production runbook for smoke tests and storage configuration for retention.