You're right. It was a simple cause after all. I think the Perc6 is
malfunctionning in some way. I have now errors I hadn't yesterday
(thousands of them). And yes the previous jobs were terminated OK, like
in your case. Bacula didn't notice anything until the fs really broke
apart. It just showed me what it could show: the sizes didn't match.
That's a good point, all in all. It is really solid software because it
didn't crash and tried to do its job until it was clear there was
something terribly wrong.
Thanks again, everybody. :-)
JF
Le 31/01/2013 19:34, Bill Arlofski a écrit :
> On 01/31/13 10:09, Martin Simmons wrote:
>>>>>>> On Thu, 31 Jan 2013 11:27:07 +0100, Jean-François Leroux said:
>>> The previous entry for this backup shows this:
>>>
>>> 29-Jan 04:35 unisson.private.tice.ac-orlea JobId 12110: Start Backup
>>> JobId 12110, Job=Backup-MyClient.2013-01-29_04.35.00_41
>>> 29-Jan 04:35 MyBaculaDir.MyDomain.com JobId 12110: Using Device
>>> "MyClient-Device"
>>> 29-Jan 04:35 MyBaculaSD.MyDomain.com JobId 12110: Volume
>>> "MyClient-Inc-0205" previously written, moving to end of data.
>>> 29-Jan 04:35 MyBaculaSD.MyDomain.com JobId 12110: Ready to append to end
>>> of Volume "MyClient-Inc-0205" size=228201991
>>> 29-Jan 04:35 MyBaculaDir.MyDomain.com JobId 12110: Max Volume jobs
>>> exceeded. Marking Volume "MyCLient-Inc-0205" as Used.
>> Can you find the log from the previous use of the volume
>> MyBaculaCLient-Inc-0064, which is the one that got the error in your previous
>> post?
>>
>> __Martin
> Hi Martin, Hi Ryan,
>
> I suspect he will see what I was seeing.
>
> Namely, the previous job that used the volume would complete OK (no errors, no
> warnings) and everything looked perfectly fine.
>
> But behind the scenes due to memory problems, SD threads were crashing and I
> didn't know it, and then the next job to use the volume would report it as
> having a size mismatch, and/or mark it in error. Heck, even when I did
> noticed the SD threads crashing, it still made no sense because the jobs were
> completing OK and there were no errors or complaints from Bacula.
>
> Because of the multitude of non-related issues all occurring at the same time,
> it was a real pita to figure out the whole problem but in the end the causes
> were pretty simple and I am glad it is back to normal again. :)
>
> Hope Jean-François has similar results. :)
>
> --
> Bill Arlofski
> Reverse Polarity, LLC
>
>
> ------------------------------------------------------------------------------
> Everyone hates slow websites. So do we.
> Make your web apps faster with AppDynamics
> Download AppDynamics Lite for free today:
> http://p.sf.net/sfu/appdyn_d2d_jan
> _______________________________________________
> Bacula-users mailing list
> Bacula-users AT lists.sourceforge DOT net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_jan
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
