On Tue, Sep 27, 2011 at 10:30 AM, Boudewijn Ector
<boudewijn AT boudewijnector DOT nl> wrote:
>
>> Do you have "Maximum Concurrent Jobs" set in the Director and storage
>> sections in bacula-dir.conf?
>
> I just added it to the storage section; it seems to have been removed
> somehow.
>>> Can someone please explain to me why bacula still is not able to run
>>> concurrent Jobs? Do I have to create a storage for each client (for
>>> instance)? And what's the reason for having to do so?
>>>
>> Only 1 volume and thus pool can be loaded in a storage device at a
>> time so if you have several pools that you want to run backups on you
>> need more than 1 storage device. For disk based backups, I highly
>> recommend using the bacula virtual autochanger.
>>
>> http://sourceforge.net/projects/vchanger/
>>
>> This will greatly simplify the setup of multiple pools, devices and
>> concurrency. Just send all jobs to the virtual autochanger resource
>> and let bacula handle the devices.
>>
> Is there any configuration required for doing so? the autochanger seems
> (to me) to be fairly complex.
>
Yes it has a small configuration file. And probably 10 to 20
additional configuration lines in your bacula-sd.conf
For me it was like 30 minutes to read the vchanger documentation and
to implement then just let it do its work. This in my opinion is much
easier than creating a separate storage device for each pool or
juggling storage devices to obtain concurrency.

>
>> Software compression is a very heavy CPU usage process on the FD and
>> will certainly slow down your backups.
>
> When having a look at the FD hosts, bacula-fd doesn't really pop up when
> running 'top', nor does the system's load increase a lot (these machines
> are also quite over dimensioned for their purpose). But that's of later
> concern, I'd be very very happy if I would be able to get concurrent
> jobs to work first.
>

Then you need do some debugging to track down the bottleneck. It can
also be the database or a number of other reasons. Also are you
talking about Full backups? I ask that because usually Incrementals
and Differentials spend more time looking for files then actually
backing up data so these tend to have low rates however since they
only backup a fraction of the Full backup they still finish much
faster.

John

------------------------------------------------------------------------------
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