Bacula-users

Re: [Bacula-users] Serveral issues potential beeing bugs

2008-07-08 03:55:27
Subject: Re: [Bacula-users] Serveral issues potential beeing bugs
From: Arno Lehmann <al AT its-lehmann DOT de>
To: bacula-users AT lists.sourceforge DOT net
Date: Tue, 08 Jul 2008 09:55:15 +0200
Hi,

07.07.2008 17:48, S.Lehmann8 AT DeutschePost DOT de wrote:
> Hi,
> 
> I have some things witch could be possible bugs.
> 
> First:
> Configured are one virual autochanger of type file and one physical 
> autochanger of type tape. There is one pool for every changer. 
> I use bweb, so i can see some informations at a glance, for example the 
> in-changer flag. 
> 
> Sometimes i get an error when changing the volume in the virtual autochanger. 
> At this time, the in-changer-flag from one physical medium (tape) ist unset. 
> After a while, many medias have an unset in-changer flag.
> 
> Here is the error message from the job, but i don't belive this is the 
> originaly reason:
> 
> 2008-07-07 15:14:59 dss-bacula-dir JobId 9485: Using Device "SL500-1-Drive-1"
> 2008-07-07 15:14:59 dss-bacula-sd JobId 9485: 3301 Issuing autochanger 
> "loaded? drive 0" command.
> 2008-07-07 15:15:00 dss-bacula-sd JobId 9485: 3302 Autochanger "loaded? drive 
> 0", result is Slot 17.
> 2008-07-07 15:15:00 dss-bacula-sd JobId 9485: 3307 Issuing autochanger 
> "unload slot 17, drive 0" command.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: Warning: Volume "00115" is in 
> use by device "FileStorage1" (/stage0/drive1)
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3301 Issuing autochanger 
> "loaded? drive 0" command.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3302 Autochanger "loaded? drive 
> 0", result: nothing loaded.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: Warning: acquire.c:207 Read 
> open device "FileStorage" (/stage0/drive0) Volume "00115" failed: 
> ERR=dev.c:491 Could not open: /stage0/drive0, ERR=No such file or directory
> 
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3301 Issuing autochanger 
> "loaded? drive 0" command.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3302 Autochanger "loaded? drive 
> 0", result: nothing loaded.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3304 Issuing autochanger "load 
> slot 15, drive 0" command.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3305 Autochanger "load slot 15, 
> drive 0", status is OK.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3301 Issuing autochanger 
> "loaded? drive 0" command.
> 2008-07-07 15:18:00 dss-bacula-sd JobId 9485: 3302 Autochanger "loaded? drive 
> 0", result is Slot 15.
> 2008-07-07 15:18:01 dss-bacula-sd JobId 9485: Ready to read from volume 
> "00115" on device "FileStorage" (/stage0/drive0).
> 2008-07-07 15:18:01 dss-bacula-sd JobId 9485: Forward spacing Volume "00115" 
> to file:block 0:793368764.
> 
> 
> As you can see, loading the required volume is no problem. I think, the error 
> message is NOT the problem of this behavior. I believe i have seen this 
> in-changer-flag setting sometimes when a media is changed. The bug is, that 
> the flag is set to the wrong media. It is set to an physical media from the 
> physical changer, which is an other pool. But the physical Volume has the 
> same slot-Number!

That would be a bug worth reporting.

> 
> Second:
> Migration jobs migrate jobs twice. In our environment every three hours a 
> migration job is beeing triggert. After that three hours it is possible, that 
> not all migration jobs are finished, so some jobs have not the status "M" for 
> migrated jobs. A new migration job is starting and those not completed jobs 
> will be queued again. Thats ok, because at this momenten the flag "M" for 
> migration is not beeing set to the job, as i wrote. Now the jobs from the 
> first migration-job will be executed and the status will be set. When the 
> second migration job for the same jobid is on the line, it is beeing migrated 
> again, although the job-status for migration is set!
> 
> Better would be to look at the beginning of a job which job-status is set.

Well, that's just the way Bacula works.

Others have stated they consider that a bug or a design deficiency. 
Anyway, Kern is unlikely to change it as it would cahnge Bacula's 
behavior for a given configuration.

Instead, a new feature for this will be implemented some day - if you 
look for "Job proximity" or "Duplicate Jobs" in the archive of -devel, 
you will find at least one of the threads discussing that.

Arno

> 
> If you thing these tings are bugs i would report it with as many informations 
> as i can provide.
> 
> 
> 
> Mit freundlichem Grüßen
> 
> Sebastian Lehmann
> 

-- 
Arno Lehmann
IT-Service Lehmann
www.its-lehmann.de

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

<Prev in Thread] Current Thread [Next in Thread>