On 10/6/2011 8:35 AM, Ben Walton wrote:
> Excerpts from shouldbe q931's message of Thu Oct 06 07:10:54 -0400 2011:
>
>> With most active/passive paired servers that I have implemented,
>> backup is run on the "passive" server and is the primary recovery
>> source.
> Yes, that's where I was headed. I guess that I probably wasn't clear
> that I was asking more for bacula best practices with such a setup
> than overall practices (eg: policy). Sorry for not being clear.
>
> Is it even sane to point bacula at the shared ip and treat the two
> systems as three clients?
>
>> You mention having a shared IP, for "clustered" systems, I would
>> probably look to have a more frequent sync than daily unless you can
>> "replace" the daily "changes" via a different method.
> This won't be a clustered system as the failover is meant for rapid
> recovery not "no data loss." (We are running drbd with corosync for
> other situations, but that doesn't fit in this case.) The shared IP
> in this instance is just so that a name can move between the systems
> if required. The sync interval could be more frequent, but the point
> is that it's not real-time.
Why not add an ethernet alias interface to the passive system and back
up only the passive system through the alias IP address?
------------------------------------------------------------------------------
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. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|