Backup & Restore
Overview
Section titled “Overview”The appliance-only Backup & Restore page is an inspection aid, not an operational backup or restore mechanism.
How it works
Section titled “How it works”The current Edge routing does not expose the page’s backup-export endpoints or its Registry/Auth SQLite inspection endpoints. Do not rely on the visible Create Backup controls: they are not connected to an operational backend.
The working inspection path reads a document-store JSON backup locally in the browser and summarizes its metadata. It performs no writes and does not execute a restore. Deployment backups and recovery remain operator-run procedures outside this page.

Step-by-step
Section titled “Step-by-step”- Before a risky change, ask an authorized operator to create a deployment backup using the deployment recovery runbook.
- On an appliance deployment, open Admin → Account → Backup & Restore only if you already have a document-store JSON backup to inspect.
- In Dry-run Restore, select that JSON file.
- Review its reported environment, tenants, collections, and record counts.
- If recovery is required, hand the file to an authorized operator and follow the deployment-specific recovery runbook. The UI does not apply it.
Common mistakes
Section titled “Common mistakes”Do not assume Version History is a single-dashboard recovery tool. Its restore operation rewinds every later checkpoint in the selected environment and tenant and can stop after partial progress. Inspect the tenant-wide impact before using it.
Do not mistake a successful dry-run inspection for a completed restore. Inspection validates and summarizes files; an operator-run recovery is a separate procedure.
Do not treat a successful browser download prompt as proof of a backup. The export routes are unavailable until they are deliberately wired, authorized, and end-to-end tested.