On this page
Package one organization's data into an archive and reload it later, behind a typed confirmation and an automatic safety snapshot.
Who can do this: an owner or admin. Super admins use the twin page under the organization's record, which adds schedules and off-platform storage.
Before you start#
- The platform's object storage is configured; otherwise the page says so instead of failing.
- This is not disaster recovery: table-scoped, one organization at a time, no point-in-time restore.
- Scheduled backups: status.
Steps#
- Open Settings > Backup & restore ("Cài đặt > Sao lưu & phục hồi") and pick a reviewed scope: Full organization, Operational data, Third-party market data, First-party and third-party data, Customer discussions, or Custom data groups.
- Check the data groups, decide on Include sync state (off by default), then press Create backup and confirm.
- In Recent backups, check the completed job shows its size. Download it, or Delete one you no longer need.
- To restore, choose a completed backup and the groups to overwrite, type the organization's slug, then press Start restore.
- A safety snapshot is written first, in the same transaction, and cannot be skipped. Afterwards it is listed in Recent backups.
If it fails#
| Symptom | Cause | Fix |
|---|---|---|
| "Storage not configured" | The platform's object storage is not set up | Ask the platform operator. |
| A restore request is refused | That mode is not available to your role | Only a super admin restores into a different organization. |
| "The 10-backup limit is reached" | Ten completed manual backups already exist | Delete an old one first. |
| A backup cannot be deleted | A restore record still references it | It stays until no restore depends on it. |
| Run failed, "fix that, then request the run again" | The staging disk ran out of space | Permanent for that run; free space, then request it again. |
| Run failed, retrying | A transient error with a retry left | One automatic retry after 30 seconds, then marked failed. |