Relax-and-Recover (ReaR in short) - Linux bare metal backup/restore tool

Mar 24, 2011
I only heard about this a couple of days ago. I gave it a go and it works so far.
I have a use case for a bare-metal backup and recovery tool, that has enough flexibility to allow me to use it for a 1:n recovery tool also.
I have a couple of fully hardened RHEL7.9 HP Z840 installed, and several Z840 that have RHEL6.10 installed I need to upgrade.
In-place upgrade not possible because / and /usr are on separate partitions on RHEL6.10 and that is not possible in RHEL7.x.
USB media are not allowed in the area.

The tool comes with the base channel in RHEL6.10 and RHEL7.9. It's all written in bash scripts.
Supposedly robust enough so you can put it into a weekly cron job.
There's several different backup methods. One that I'm using for now is backing up to an NFS mount. It constructs a ~600MB .iso
to boot bare-metal recovery from, and a .tgz GNU tar backup of the system.
It may just go seamlessly when doing bare-metal recovery of the same host, I don't know.
I'm using it to do a RHEL6.10 to RHEL7.9 upgrade of like-hardware hosts. It recognizes it's not on the identical host and
stops so I can put in an IP for the NIC on the new host, then run the restore.

It's a faithful recovery of the original host. My original host is Kerberized and also uses RSA tokens for logins. I have to
clear out all the Kerberos caches and machine keys, and the machine securid secret, and recreate those for the new host.
Zero out all the logs/audit logs.

I'm wondering if I can use it to make / and /usr share the same partition on a couple of RHEL6.10 servers,
so I can do an in-place upgrade some time in the future. Not sure yet.

It's supposed to work on AIX also (IBM supports it), which will be nice because I have LPARS I need some
disaster recovery procedure for.

More info from Redhat:

The website for the tool is:
The "about" page follows:

About Relax-and-Recover
Relax-and-Recover has a few interesting characteristics you may find useful when assessing it as a Disaster Recovery solution:

  • Modular design, written in Bash
    • easy to extend with custom functionality
    • targeted at sysadmins foremost
  • Set up and forget nature
    • designed to be easy to setup
    • designed to require no maintenance (e.g. cron integration, nagios monitoring)
  • Recovery image based on original distribution with original tools
    • recovery process remains compatible with original system and applications
    • hardware support is guaranteed
  • Two-step recovery, with optional guided menus
    • disaster recovery process targeted at operational teams
    • migration process offers flexibility and control
  • Bare metal recovery on dissimilar hardware
    • support for physical-to-virtual (P2V), virtual-to-physical (V2P)
    • support for physical-to-physical (P2P) and virtual-to-virtual (V2V)
    • various virtualization technologies supported (KVM, Xen, VMware)
  • Support for various integrated boot media types, incl.
    • ISO
    • USB
    • eSATA
    • OBDR/bootable tape
    • PXE
  • Support for various transport methods, incl.
    • HTTP
    • HTTPS
    • FTP
    • SFTP
    • NFS
    • CIFS (SMB)
  • Extensive disk layout implementation, incl.
    • HWRAID (HP SmartArray)
    • SWRAID
    • LVM
    • multipathing
    • DRBD
    • iSCSI
    • LUKS (encrypted partitions and filesystems)
  • Supports various 3rd party backup technologies, incl.
  • Supports various internal backup methods
  • Two phase disk layout recovery, allows reconfiguration before recovery, e.g.
    • migrations from e.g. SWRAID to HWRAID, or unencryped to encrypted partitions
    • HWRAID reconfigurations
    • migration from partitions to LVM
  • Various techniques to help troubleshooting
    • structured log files and guided menus
    • log files are moved to recovery image, and to recovered system (available in every step for debugging)
    • advanced debugging options to help trace scripts or develop new functionality
  • Integration with monitoring (examples for Nagios/Opsview)
  • Integration with scheduler (e.g. let cron recreate and transfer your images upon disk layout changes)
  • Various best practices to assist recovery
    • integrates with local bootloader (in case it is still possible, you can restore from local disk through Grub)
    • automatic network and ssh configuration (for remote recovery)
    • automatic serial console support (useful for recovery through iLO or KVM serial console)
    • shell history-stuffing (stuff shell history with useful commands at every step)
    • automatic recovery when possible, guided recovery when needed
Was this message intended for a computer forum? I like the idea of learning about computers, but this is WAY over my head. And I intend to try Linux someday soon..
That is very cool! Not directly applicable to my PACS applications as the server environment (RHEL) outside the PACS itself isn't important, it's just a generic install, but I could see this making sense in a more complex role.

Thank you very much for the share!
I am not in/into "corporate world"...

But user said: And I intend to try Linux someday soon..

Thats why I have posted it
Last edited: