Bacula-users

Re: [Bacula-users] pruning of virtual full jobs

2014-12-16 12:59:48
Subject: Re: [Bacula-users] pruning of virtual full jobs
From: Josh Fisher <jfisher AT pvct DOT com>
To: bacula-users AT lists.sourceforge DOT net
Date: Tue, 16 Dec 2014 12:55:48 -0500
On 12/11/2014 12:11 PM, Thomas Lohman wrote:
> This is probably a question for Kern or perhaps should be better posted
> to bacula-devel but I'll send it here since others may have experienced
> or have comments on this.
>
> Assume you are running Virtual Fulls every x days (aka the Max Full
> Interval for Virtual Fulls) and also have retention periods for
> clients/volumes set.  When a client that comes and goes is ready for a
> new Virtual Full, it's possible that there have been no new
> Incremental/Differential backups since the last Virtual Full.  So, it
> simply makes a new "copy" of the last Virtual Full which makes sense.
> When you then run a prune of that client, it will look at the JobTDate
> of the Virtual Full job and see the date of the original last "real"
> backup for that client and depending on the retention defined will
> delete the job information which then leads to an error on it's next
> backup attempt.  At this point, you have to get the client in and do a
> new Full for that job.
>
> The issue really seems to be whether or not for Virtual Fulls, pruning
> should use the real job termination time and not the job termination
> time that gets dragged forward from the last "real" backup that was
> done.  It seems to me that it should but I can see an argument the other
> way as well since the actual data you're storing has aged past your
> retention periods.

It should always be the last "real" backup time, since that is what 
coincides with the stored data. Anything else would be misleading and 
easily confused. What I do is assign long retention times to the full 
jobs (and so virtual full) for those roaming clients. It means I have to 
manually prune old full jobs for those clients, but there aren't too 
many of them in my case, and it is better than scheduling time for a 
full job run. Those users tend to get upset if they have to stick around 
long enough for a full backup to complete on their slow laptop.

Be sure to review the RescheduleOnError and associated directives to try 
to increase the chances of getting successful incremental jobs for those 
roaming clients.


------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

<Prev in Thread] Current Thread [Next in Thread>