Skip to content

Restore files from a recovery point

View as Markdown

A restore copies files and folders from one of the node’s recovery points back onto a node. You pick exactly what to bring back, and the rest of the node stays as it is. You can restore to the source node, the one the backup came from, or to another of your nodes.

The files go into a folder you choose on the node: a new folder MyAccount names, or a folder you type. You then move them to where they belong.

  • The source node is running. You can restore its files only while it runs. For a kept backup of a deleted node, see Keep backups after deleting a node.
  • The node you restore to has enough free disk space for the files. A restore fails when the disk runs out of space.
  • If you restore to another node, its first backup has finished: its Backup Status reads Available. The source node is always listed in Restore on Node.
  • For an encrypted backup, you have its passphrase; you need it to browse or restore the backup.
  1. In Manage CDP Backups, click the node and open its Backup Recovery Points tab. Each recovery point is listed with its ID and the date and time it was created, newest first.
  2. Find the recovery point whose files you want back. Each one holds all the node’s backed-up data at that moment. You need only that one.
  3. Open its Actions and click Browse or Restore, or click its date. Browse Recovery Point opens.
  4. (Encrypted backup only) Enter the passphrase in Passphrase and click Submit. A wrong passphrase shows “Passphrase provided is incorrect. Please provide correct passphrase to proceed.”

Browse Recovery Point opens at /, the top of the node’s disk.

  1. Click a folder to open it, or type a path in Location and click Go to jump there.
  2. Check the files and folders you want.
  3. Click Restore Selected.

Only the items checked in the folder on screen are restored. To bring back files from two different folders, restore each folder in turn.

Check the node and the folder before you click Restore. Once a restore starts, it runs to the end.

  1. In Restore on Node, choose the node to restore to. No node is chosen for you. The source node is marked (Source Node). The list also has your other nodes in the same region whose first backup has finished, with the same kind of operating system: a Linux backup restores to a Linux node, a Windows backup to a Windows node.

  2. Under Location, choose where on that node the files go:

    • Default writes them into a new folder under /e2e-backup-restore/ on Linux, or c:\e2e-backup-restore\ on Windows, named after the date and time of the restore in UTC. Path shows the folder. Choose this for most restores: it leaves everything already on the node as it is.

    • Alternate writes them into a folder you type in Path, as an absolute path such as /home/restore, or D:\restore on Windows.

      Files already in that folder stay as they are. If a file with the same name is already there, the restore skips it. MyAccount still warns “Proceed with caution, as making changes may overwrite your existing files.” To get back every file you selected, choose an empty folder or a new one.

      Restore into a folder, never to / itself. Restoring to / copies the backed-up operating system files onto the node and can fill its disk.

      A path that is not absolute gets “Please provide correct restore path location.”

  3. Click Restore. With Default, the restore starts at once. With Alternate, check the path shown, and click Confirm Restore.

You see “Your backup restoration is in process.” The files arrive under their own names, relative to the folder you browsed: restoring a.txt and sub while browsing /root/app gives a.txt and sub/ in the restore folder. To undo a restore, delete the restored files from the folder.

A restore takes longer the more data you select. Open the Restore Task History tab of the source node. Each restore is listed there with its start time, end time, and duration, and a restore to another node also shows that node’s name. Task Status reads Finished once it completes.

E2E emails the account owner when the restore ends: “Backup Restoration completed for your Node”, or “Backup Restoration failed for your Node” if it fails. The “failed” email also arrives for a restore that reads Finished but skipped a file that was already there; see The restore email says it failed, but the restore reads Finished.

If Task Status reads Failed or Warning, check the restored files and the email E2E sends when the restore ends. See The restore failed.

On the node, copy or move the restored files from the restore folder to where they belong. To bring a database or a website back, also check file permissions, and restart its services.

Move every file you need out of /e2e-backup-restore or c:\e2e-backup-restore. After a restore, the source node’s backups leave that folder out. On Linux, only root can open the folder a Default restore creates.

You now have the files back on the node. To keep this recovery point from being merged away, see Lock a recovery point. To restore the files of a node you have deleted, see Keep backups after deleting a node. If the restore does not start or fails, see Troubleshooting CDP Backup.