We use simply a pool for each server/Type of jobs here
with 10 servers it give us a
30 pools
server1D ( days )
server1W ( weeks )
server1M ( months )
...
Each pools has it's own rules ( recycling etc ..) and number of media,
We launch all them concurrently and everything work as expected (Sure Dir and
Sd works a lot :-)
The other way you can do things are by using spooling which has it's pro/co.
see manual for that.
Kevin Keane wrote:
> Ulrich Leodolter wrote:
>> On Sat, 2008-11-15 at 01:23 -0800, Kevin Keane wrote:
>>
>>> I'm using Bacula to back up 8 servers to a pool on a hard disk. To make
>>> management somewhat easier, I configured bacula to create a new volume
>>> (i.e., file) for each job. In the past, I only allowed one backup job to
>>> run at a time (Maximum Concurrent Jobs = 1) and everything worked
>>> beautifully.
>>>
>>> Today, I tried to speed up my backups by allowing multiple concurrent
>>> backup jobs. Sure enough, the backups completed much faster - but
>>> apparently multiple jobs now ended up in the same volume, despite the
>>> Max Volume Jobs setting.
>>>
>>> Is this a bug? If so, is it something that has already been reported
>>> and/or fixed? I'm using 2.4.2.
>>>
>>>
>> Hi,
>>
>> It is documented, read the "Maximum Volume Jobs" description at:
>> http://www.bacula.org/en/rel-manual/Configuring_Director.html#SECTION0014150000000000000000
>>
>> Ulrich
>>
> Ah! Thanks. In other words, there is a race condition. Is there a
> workaround (other than foregoing concurrent jobs)?
>
> Also, can you add a fix for this to the wishlist for bacula 3?
>
> BTW, it seems that the actual behavior may be even worse than
> documented. The documentation says that jobs may fail - but mine all
> succeeded. According to the log file, they probably clobbered the
> volume. Several of the jobs not only report that they are going to use
> the same volume, but also that they are *overwriting* the same volume. I
> see the following in the log for five separate backup jobs
> (interestingly, only one says "wrote label..." - the other ones just
> mark it as purged again and recycle it).
>
> (in only one job)
>
> 14-Nov 23:05 xxx-dir JobId 568: There are no more Jobs associated with Volume
> "bacula0420". Marking it purged.
> 14-Nov 23:05 xxx-dir JobId 568: All records pruned from Volume "bacula0420";
> marking it "Purged"
> 14-Nov 23:05 xxx-dir JobId 568: Recycled volume "bacula0420"
> 14-Nov 23:05 xxx-dir JobId 568: Using Device "eSATADisk3"
> 14-Nov 23:05 xxx-sd JobId 568: Labeled new Volume "bacula0420" on device
> "eSATADisk3" (/misc/BACKUP3).
> 14-Nov 23:05 xxx-sd JobId 568: Wrote label to prelabeled Volume "bacula0420"
> on device "eSATADisk3" (/misc/BACKUP3)
> 14-Nov 23:05 xxx-dir JobId 568: Max Volume jobs exceeded. Marking Volume
> "bacula0420" as Used.
>
> (in all other jobs)
> 14-Nov 23:05 xxx-dir JobId 573: There are no more Jobs associated with Volume
> "bacula0420". Marking it purged.
> 14-Nov 23:05 xxx-dir JobId 573: All records pruned from Volume "bacula0420";
> marking it "Purged"
> 14-Nov 23:05 xxx-dir JobId 573: Recycled volume "bacula0420"
> 14-Nov 23:05 xxx-sd JobId 573: Recycled volume "bacula0420" on device
> "eSATADisk3" (/misc/BACKUP3), all previous data lost.
> 14-Nov 23:05 xxx-dir JobId 573: Max Volume jobs exceeded. Marking Volume
> "bacula0420" as Used.
>
> Thanks for your help!
>
> Regards,
>
--
Bruno Friedmann
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
