Backup & Recovery
This page shows all backups of your shop and lets you create a manual backup, restore an existing backup, or re-upload a previously downloaded backup file. It also shows the disaster recovery plan and the configured backup schedules.
What can I do here?
- Create a manual backup — optionally choosing specific layers, tenants and tables
- Download or restore an existing backup
- Re-upload a previously exported backup and restore from it
- Cancel a running backup
- View the disaster recovery plan with recovery targets (RTO/RPO)
- View the automatic backup schedule and storage tiers (Hot/Warm/Cold)
Step by step
- Create a manual backup: Click "Manual backup", give it a name, choose the desired layers (Platform, Tenants, Media manifest) and optionally specific tenants/tables, then start the process.
- Restore a backup: Click "Restore" next to a completed backup in the table, choose the parts to restore (database, media, configuration, themes), and confirm with the security code.
- Download a backup: Click the download icon next to the relevant backup.
- Restore from a file: Click "Upload & restore" and select a previously downloaded backup file.
- Check the schedule: Open the "Settings" tab to see the automatic schedule (full backup, incremental, weekly archive, pre-deploy snapshot) and the storage tiers.
Fields explained
| Field | Meaning | Notes/Impact |
|---|---|---|
| Backup name | Label for the backup | Freely chosen |
| Type | Kind of backup (Full, Incremental, Manual, Pre-deploy) | Determines scope and trigger |
| Size / Duration | Size and creation time of the backup | Informational only |
| Content | Which parts are included in the backup | Shown as a short description or tag list |
| Status | Current state of the backup | See table below |
| Encrypted | Marks whether the backup is stored encrypted | Increases data security |
| Restore – Database/Media/Config/Themes | Selection of which parts get restored | Unselected parts stay unchanged |
| Confirmation code | Security code you must type during a restore | Prevents accidental restores |
Values & statuses
Backup status
| Value | Plain meaning | What happens |
|---|---|---|
| Completed | Backup was created successfully | Can be downloaded or restored |
| Running | Backup is currently being created | Can be stopped via "Cancel" |
| Failed | Backup creation failed | Not usable; try again if needed |
| Scheduled | Backup is planned for a later time | Starts automatically |
| Cancelled | Backup was manually stopped | No usable result is available |
Backup type
| Value | Plain meaning |
|---|---|
| Full | Complete backup of all data |
| Incremental | Backup of only the changes since the last backup |
| Manual | Backup manually triggered by a user |
| Pre-deploy | Backup automatically created before a deployment |
Storage tiers
| Value | Plain meaning |
|---|---|
| Hot | Instant access (under 1 second) |
| Warm | Fast access (under 5 minutes) |
| Cold | Archive storage, restore can take up to 12 hours |
Frequently asked questions
Are roles, permissions and passkeys part of the backup?
Yes. A backup covers the complete users-and-permissions area: user accounts, roles, role assignments, individual permission exceptions, passkeys (password-free sign-in), password history, sign-in tokens and the session overview. On restore they are replayed in the correct order - the role first, then its assignments - so a restored staff member has the same roles and the same sign-in methods as before the backup. Backup files created before 2026-08-12 do not contain that information; in that case please create a new backup. If you pick individual tables in a manual backup, the interface warns you about missing dependencies (for example role assignments without their roles) - follow those warnings, otherwise the backup can only be restored in part.
Can a backup be reported as "Completed" even though something is missing?
Not any more. If an expected table cannot be read while the backup is being created, the run ends with the status Failed, names the affected table and stores no file at all. The previous behaviour - store the file, report "Completed", mention the problem only in the log - was the more dangerous one: an incomplete file looks like a backup, and you only find out when you restore it. A failed run deliberately cannot be restored; start it again once the stated cause is fixed.
Can a restore be undone?
No — a restore is a destructive operation: all data created since the selected backup is lost. Because of this, a snapshot of the current state is automatically created before the restore begins.
What do RTO/RPO mean in the disaster recovery plan?
RTO (Recovery Time Objective) indicates how quickly a system should be available again after an outage. RPO (Recovery Point Objective) indicates the maximum tolerated data loss (e.g. "under 1 hour").
Are the security toggles (Encryption, Geo-redundancy, Failure notification) in the "Settings" tab configurable yet?
These toggles currently only display status and are labeled "Coming soon" — direct editing through the interface isn't available yet.