Bare Metal Restore (ReaR)

Emergency server recovery

On the server list simply click the rear icon to start the recovery:

Step one: select rescue point

Step two: download rescue image and boot server in rescue mode

Next, the system will check access to the storage location for the selected job and check if a ssh key is installed. If there is no ssh key present, it will prompt for ssh credentials. Afterwards, the system will give a link to download the rescue image:


Use the first option to restore to the same hardware. If you need to move (clone/migrate) to different hardware, select the second option and define your new network data. See migration chapter for more info.

You need burn the provided image to CD, DVD or USB stick and boot the server from it. Please note that by default it will still try to boot from its local data drive so you need manually select "automatic recovery"


!!! Important !!! The system will automatically reformat your disk and destroy all data (if there was any) according to the disk layout in the selected backup job.

Wait for while you'll see prompt to back to bacula4 GUI:

Step three: checking server status

System checks access to server and ensures that it's booted from the recovery disk. Once you see green status, click next:

Step four: Restore data

The system will proceed with restoring your files and display a restore log. File size errors can generally be safely ignored. When the restore is done it will ask you to finish:


Within a few minutes it will install a boot loader and on server console ask you to reboot the server or check the logs:



Just type "3" and press enter.

Step five: Restore MySQL

If you enabled MySQL or Percona Tools backups, you'll be prompted for 5th step for the restore of MySQL. If you did not receive errors prior on the mysql data file restores, you may be able to skip this step, however, it is generally required.

Example of dump recovery:

Example of Percona recovery:

Warning: Percona recovery works only for default mysql data folder location: /var/lib/mysql. For other installation paths, you will need to restore mysql manually via regular recovery procedure.

Server migration

Step one: prepare ISO

You can use the default iso and redefine network setting by interrupting the automatic boot process and give some extra kernel options (with tab button):

ip=192.268.1.100 nm=255.255.255.0 gw=192.168.1.1 netdev=eth0

Or you can simply select custom iso download and define network configuration for new hardware.

Step two: restore server with Relax and Recover

Next, follow the aforementioned recovery instructions, except be sure to specify the new IP when the server status check fails.



Lastly, set the new IP of destination server and click try again button. After checking access you'll able continue restore and get complete copy of your server.

ReaR under onapp

First, boot the vm from rescue iso. Navigate to onapp -> templates -> ISO list -> upload iso:


Next you can boot that VM into the ISO uploaded prior.



Note that there is a 30 sec timeout to select rescue mode so go to console and select boot option as soon as it available:



In some instances onapp changes device names e.g. from /dev/vda to /dev/sda. In such case you'll see a warning like:



If you select #5 to continue booting the vm, it will not boot after recovery. Therefore, you will see boot troubles after recovery like:



To resolve this issue, boot configurations will need to be modified. This will be common if migrating between physical hardware or virtualization platforms. Users may either go to single user mode and fix grub / fstab or boot into the onapp recovery mode:



Log into recovery console and mount your disk:



Now you need fix /mnt/etc/grub and /mnt/etc/fstab. Examples:
From (near root=):


To:


From:


To ( rename sda to vda ):


Now you just run "poweroff" in console and start vm from onapp it will boot no errors.

Bare Metal Restore for Windows (bareos only)

Step 1 Select rescue point, same way as for REAR:


Next, download your ISO image. Every server will have its own ISO image, regardless of recovery date.

There are options to either select the default ISO and define the network settings after boot or in the GUI prior:

If you select the second option, you need to redefine network settings and the system will regenerate the ISO so these parameters will not need to be entered in the recovery environment.


Once networking is established, you will need to navigate back to the Clusterlogics gui. If you need to run network setup again you can type PENETWORK in cmd console.



Next, the system checks if it able to connect to the rescue environment. If you use a temporary IP for the migration, you'll be prompted to update it here, however do not forget to update it after restore.



Now you will need to restore your partitions. If you continue with this process, you'll destroy all data on your local system disk to prepare for the restoration. Clusterlogics will review the restoration point selected and generate a diskpart script to restore your disk with similar partitioning. In case if you have several disks (e.g. more than the original machine disk amount), you can select or change disk mapping (e.g. select disks that have more space). Lastly, the system will generate a disk part script that you should review and update it if necessary.

If there are no errors, you'll see the disk part message and be able to continue:

The last step: restore data


Once your data is restored, the system will install the boot loader and ask you to remove the boot media and reboot the server: