On Nov 11, 2008, at 2:32 PM, Jason Dixon wrote:
> We have a new Bacula server (2.4.2 on Solaris 10 x86) that runs fine
> for
> most backup jobs. However, we've encountered a particular job that
> hangs indefinitely with the status "Dir inserting attributes". It's
> important to note that all of our other jobs complete successfully
> except this one.
What others have said confirmed what I thought. I too think this job
is not hanging. It is inserting.
select * from pg_stat_activity
The above will probably show the INSERT below.
Using systems tools such as top, vmstat, iostat, and ps, you should
be able to prove that progress is being made.
>
>
> Running Jobs:
> JobId Level Name Status
> ======================================================================
> 90 Full Unix_zimbra.2008-11-10_15.26.10 Dir inserting Attributes
> ====
>
> This particular job is for our Zimbra server (2.4.3 on Linux x86_64).
> The job is predictably large (275GB, 15M files). I asked our
> PostgreSQL
> DBA to review the database and he found a hung statement (backslashes
> are mine):
>
> INSERT INTO File (FileIndex, JobId, PathId, FilenameId, LStat, \
> MD5)SELECT batch.FileIndex, batch.JobId, Path.PathId, \
> Filename.FilenameId,batch.LStat, batch.MD5 FROM batch JOIN Path ON \
> (batch.Path = Path.Path) JOIN Filename ON (batch.Name = Filename.Name)
>
> At this point, other jobs (except Catalog Backup) will continue to run
> and complete successfully. But to get the Catalog Backup running, I
> have to cancel the zimbra job, restart bacula-dir, and kick off the
> Catalog Backup job. Otherwise, the Catalog jobs will just sit there
> waiting to execute. I've let these jobs sit for days in this state,
> so
> it's not like I'm being impatient. :)
Days? Oh. Hmmm. Perhaps you need to tune postgresql.conf a bit more.
What are the hardware specs for the database server?
>
>
> One other note. If the DBA kills that query in the database, the job
> completes. Although the director reports an Error, it updates the
> catalog successfully. If I cancel the job inside bconsole, nothing
> else gets updated in the catalog, presumably due to a database lock.
> The first job below (jobid 90) was one I cancelled; the next (jobid
> 104) was where the database query was killed.
>
> *llist jobid=90
> No results to list.
>
> *llist jobid=104
> jobid: 104
> job: Unix_zimbra.2008-11-11_11.19.05
> name: Unix_zimbra
> purgedfiles: 0
> type: B
> level: F
> clientid: 2
> name: zimbra
> jobstatus: f
> schedtime: 2008-11-11 11:19:29
> starttime: 2008-11-11 11:19:33
> endtime: 2008-11-11 17:10:42
> realendtime: 2008-11-11 17:10:42
> jobtdate: 1,226,441,442
> volsessionid: 16
> volsessiontime: 1,226,346,967
> jobfiles: 15,020,757
> joberrors: 1
> jobmissingfiles: 0
> poolid: 1
> poolname: Default
> priorjobid: 0
> filesetid: 2
> fileset: Zimbra Set
--
Dan Langille
http://langille.org/
-------------------------------------------------------------------------
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 AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
