The native backup engine — incremental, content-addressed, manifest-per-restore-point, writing to an S3-compatible store or a local directory — is part of the Community edition. The Pro edition extends it with audit-chain protection, restore drills, and a full disaster-recovery flow.

What Pro addsCopy Anchor Link

Audit delta chainCopy Anchor Link

Community backups snapshot the whole data root. Pro registers additional backup contributors:

• Audit deltas — only audit rows newer than the previous restore point (seq > cursor) are exported as gzip JSONL; the first run takes a VACUUM INTO baseline. When cumulative deltas exceed 50% of the baseline, the chain rebases automatically (POST /api/v1/backup/rebase).
• Integrity & timestamp snapshots — integrity.db and anchor.db are uploaded as full snapshots, skipped when their hash is unchanged.

Restore drillsCopy Anchor Link

A drill runs a verify-only restore against the configured target — S3 or local — every object is checked against its manifest, nothing is written, and a DrillReport (passed, objects checked, bytes checked) is persisted for the WebUI — your periodic, auditable proof that recovery works.

The same verification is available offline, from the CLI, without a running server:

./bichon-admin restore --config restore-config.json --bichon-restore-verify-only

That is the command to reach for on a recovery box before committing to a restore: it points at the same target the real restore would read, and tells you whether the backup can actually be rebuilt. --into is not required, since nothing is written. See CLI Tools.

Audit replayer on restoreCopy Anchor Link

The Pro flavor of bichon-admin replays the backup's audit delta chain back into a live audit.db after restore, so the hash chain continues unbroken across a disaster.

Disaster recovery ordering: restore runs before the license preflight, so a DR replacement machine needs no license to recover data.

Backup API (all editions)Copy Anchor Link