Thanks to everyone for replies.

I decided to go for Bacula-only solution though.

So I created 2 jobs: one for users' data and another one for the rest of 
the Zimbra:

=== Job 1 ===
/opt/zimbra/store
/opt/zimbra/index
/opt/zimbra/mysql/bin/mysqldump
/opt/zimbra/libexec/zmslapcat
/opt/zimbra/libexec/zmslapcat -c

=== Job 2 ===
/opt/zimbra
without:
/opt/zimbra/store
/opt/zimbra/index


I'm aware that I risk the server being slightly out-of-sync for the time 
period between dumps and file-based backups of store and index. What 
minimize the time period are LVM-snapshots of store and index.

-- 
Silver

On 25.11.2011 16:51, Silver Salonen wrote:
> Hi.
>
> Is anyone backing up Zimbra on-the-fly? I don't think taking server
> offline for pure file-based copy is a modern method of doing things.
> Neither do I want to use zmbackup, because as I understand, that dumps
> all the mailboxes (which are on disk anyway) to separate files which
> would just waste so much space.
>
> I'm currently backing up the whole /opt/zimbra and I've already managed
> to restore a user's calendar by restoring MySQL-files onto another
> server and starting MySQL from them. I guess I was lucky. And there's
> also possibility to restore users' .msg files to submit them into Zimbra
> server again. So I don't see much problem here..
>
> Anyway, I wonder what is there to backup anyway? Store, MySQL, LDAP, ...
> something more?

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users