[rear-users] Rear Recovery from OS Rescue Mode

Johannes Meixner jsmeix at suse.de
Fri Apr 1 11:44:31 CEST 2016


Hello,

On Mar 31 14:04 Jeffrey Huckaby wrote (excerpt):
> Can rear be ran from within a native rescue
> mode boot environment?

Hooray!
You are the first non-SUSE user who asks for such
kind of functionality.

I am very much interested in users of other Linux
distributions who also like to have such kind of
functionality.

I developed a proof-of-concept of a similar
functionality with the SUSE installation system,
see the section
"Using rear as generic installer in any installation system"
in
https://en.opensuse.org/SDB:Disaster_Recovery

You asked for "Running the rear recovery from
within a native rescue system".

My proof-of-concept is just one step further:

Instead of "dump the payload" by restoring
the files from a backup (that is what you ask for),
in my proof-of-concept I "dump the payload" by
installing RPM software packages.

Regarding what I mean with "dump the payload" see
the section
"Disaster recovery is installation (reinstalling from scratch)"
in
https://en.opensuse.org/SDB:Disaster_Recovery


> For example, if I just boot into a CentOS rescue mode
> can I then configure rear and run the recovery?

You need to distinguish between
- configure rear and run "rear mkbackup" versus
- boot a recovery system and run "rear recover".

Configuring rear and then "rear mkbackup"
happens in the original running system.

In contrast "rear recover" is reinstalling from scratch
which must run from within an installation system
which is currently the rear recovery system.

There is no real hard limitation why "rear recover"
could not also run from within any sufficiently
powerful installation system, for example any
sufficiently powerful "rescue system".

The first precondition is that an installation system
whereform "rear recover" should be run must contain
the rear bash scripts plus the rear configuration files.

Therefore the first task is how to enhance an existing
installation/rescue system so that it contains the
rear bash scripts plus the rear configuration files.

Perhaps the easiest way is to boot the existing
installation/rescue system and from within that system
download the rear bash scripts plus the rear configuration
files from an appropriate other machine in the network
(which means appropriate tools to download files must
be available in the installation/rescue system).

Things could get complicated if an existing rescue system
is not yet sufficiently powerful to be also used as an
installation system (e.g. because certain tools are
missing that are needed for system installation).

But I assume that the usual tools for disk partitioning,
creating filesystems, and boot loader installation
are available on usual rescue systems so that usual
rescue systems should be sufficiently powerful to be
also used as an installation system - at least for
"usual" system installations.

Therefore the only thing left is a tool to restore
the files from the backup. When the backup is
a tar.gz archive, I assume that 'tar' is available
on usual installation/rescue systems so that
"rear recover" should be able to run successfully
from within usual installation/rescue systems
when the backup is a tar.gz archive.

In contrast when third-party backup software was used
to make the backup of the files during "rear mkbackup",
then the installation/rescue system where "rear recover"
should run must be enhanced with the appropriate tools
to restore this third-party backup.


Kind Regards
Johannes Meixner
-- 
SUSE LINUX GmbH - GF: Felix Imendoerffer, Jane Smithard,
Graham Norton - HRB 21284 (AG Nuernberg)



More information about the rear-users mailing list