On 3/2/2012 6:02 AM, Wassim Zaarour wrote:
Hi List,
We are backing up to disks not tapes, now I need to set up a plan in
case the bacula server (director) crashes.
What to do in case the MySQL catalog is lost? Can we recover? I guess
it is easy to keep copy of configuration files to install on a new
system but what about the catalog?
What is the best plan for the server's disaster recovery?
Thanks.
I use a RunScript that runs after the last job of the night. The script
saves an sql dump, config files, and bootstrap files to a NFS share on
another server. The server hosting the NFS share is, of course, also
backed up, giving a means to restore a one day old copy should both
servers fail.
Bacula Dir and SD run in a qemu-kvm VM. The advantage is that it is
possible to send a complete Bacula server image and XML libvirt
definition along with the offsite backup media. The VM image is a ready
to run Bacula server complete with up to date catalog. Restore from a
total loss is then simply dd'ing the VM image to new hardware and
starting the VM.
I suppose it depends on the infrastructure being used. In this case,
production servers are all VMs running on a two node Pacemaker/Corosync
cluster, so the above is a good fit. Since Bacula can run on either node
at any time, a hardware failure on a single node does not require a
restore, so more attention has been given to offsite recovery. However,
the nightly scripted backup of the catalog is still needed to protect
against weird database or filesystem errors.
------------------------------------------------------------------------------
Virtualization & Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
