Bacula-users

Re: [Bacula-users] large jobs that fail

2014-11-07 08:39:23
Subject: Re: [Bacula-users] large jobs that fail
From: Bryn Hughes <linux AT nashira DOT ca>
To: bacula-users AT lists.sourceforge DOT net
Date: Fri, 07 Nov 2014 05:37:02 -0800
On 2014-11-07 04:40 AM, Jeff MacDonald wrote:
> Lets say you are doing a 500 gig job and it fails after 450 gigs.. and you 
> are using disk based backups.
>
> Do you ever bother to try to reclaim that space? I mean I delete the job, so 
> the volume can have that data recycled when the time comes, but do you ever 
> go thru the extra trouble of finding the volumes that only have data from 
> that job on , and deleting them from disk etc?
>
> jeff.
>
Are you asking if I PERSONALLY try to reclaim the space, or if Bacula 
itself will try to?

Bacula will keep the backed up data and apply the normal retention 
policies to it even if the backup doesn't complete successfully. That 
450 gigs of data on disk can still be used for a restore despite the 
backup not completing successfully; the complexity of restoring the data 
will depend on the circumstances around how it failed.  If for instance 
the director crashed and the records never got inserted into the 
database restore is going to be a lot more complex than if the client 
went offline for some reason.  However the data is still there and can 
still be read one way or the other.

On a personal level, whether I try to reclaim the space or not will 
depend on whether or not I have enough space to have that partial backup 
sitting around or not.  If it's not causing any issues then I'd be 
inclined to let it sit till it expired on its own.

Bryn



------------------------------------------------------------------------------
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users