Backups and restores
A backup is the safety net for Beam device operations. Take one before a software change or a service removal, and you have a configuration you can roll back to. A backup captures the device’s services, templates, failover groups, server definitions, and local users. Backup and restore operations run asynchronously on the device and return 202 Accepted.
Create a backup
A backup is triggered with POST against the device. Use a descriptive backupName, such as one with a date, so its purpose is clear later.
- Terminal window
The call returns 202 Accepted and the backup completes on the device over the next while. Configuration access on the device is blocked while a backup or restore is in progress.
Browse backups
List the backups in a project, and read one to get its stored file details:
A backup record carries its timestamp, the blobStorageUrl of the stored file, and a signature for integrity verification. The list endpoint supports $filter; see the Fleets API reference for the filterable fields.
Restore a backup
Restore is also a POST, with the stored file path from the backup record. It returns 202 Accepted and applies asynchronously.
Confirm the backup identity before you restore it onto a live device.
Upload an external backup
If you already hold a .tar.gz backup file, upload it and let MK.IO generate its metadata:
What goes wrong
Restoring a backup onto a different Beam server. This is not supported. A backup restores to the device it came from. See the on-device detail in Backup and restore.
-
Old backups disappearing. A device keeps up to 30 local backups and deletes the oldest when that limit is reached. Download or upload any you must retain.
-
Configuration changes blocked mid-operation. Device configuration access is locked while a backup or restore runs. Wait for it to finish.
What comes next
-
Software and updates: take a backup before changing software state.
-
Manage devices: return to the device once protection is in place.