Bacula-users

Re: [Bacula-users] virtuafull and verify

2009-11-28 03:30:31
Subject: Re: [Bacula-users] virtuafull and verify
From: "Fahrer, Julian" <julian AT fahrer DOT net>
To: <bacula-users AT lists.sourceforge DOT net>
Date: Sat, 28 Nov 2009 09:24:59 +0100
Hey guys,

after changing the SQL-Statement in the source code, the right jobs are 
slected. But to me it looks like verifying a Virtual Full doesn't work 
correctly. Here is a Job Output:

28-Nov 06:29 backup01_dir JobId 3325: Verifying against JobId=3320 
Job=nas01_zarafa.2009-11-28_01.00.00_45
28-Nov 06:29 backup01_dir JobId 3325: Bootstrap records written to 
/var/bacula/working/backup01_dir.restore.48.bsr
28-Nov 06:29 backup01_dir JobId 3325: Start Verify JobId=3325 
Level=VolumeToCatalog Job=nas01_zarafa_verify.2009-11-28_06.00.00_50
28-Nov 06:29 backup01_dir JobId 3325: Using Device "LTO2"
28-Nov 06:30 backup01_sd JobId 3325: Ready to read from volume "LTO2_Fri-0003" 
on device "LTO2" (/dev/nst0).
28-Nov 06:30 backup01_sd JobId 3325: Forward spacing Volume "LTO2_Fri-0003" to 
file:block 50:0.
28-Nov 06:55 backup01_fd JobId 3325: Fatal error: Record header file index 180 
not equal record index 202 28-Nov 06:55 backup01_sd JobId 3325: Fatal error: 
read.c:122 Error sending to File daemon. ERR=Connection reset by peer 28-Nov 
06:55 backup01_sd JobId 3325: Job nas01_zarafa_verify.2009-11-28_06.00.00_50 
marked to be canceled.
28-Nov 06:55 backup01_sd JobId 3325: Fatal error: fd_cmds.c:168 Command error 
with FD, hanging up.
28-Nov 06:55 backup01_sd JobId 3325: Error: bsock.c:374 Write error sending 33 
bytes to client:192.168.10.76:36643: ERR=Connection reset by peer 28-Nov 06:55 
backup01_dir JobId 3325: Warning: The following files are in the Catalog but 
not on the Volume(s):
28-Nov 06:55 backup01_dir JobId 3325:       /mnt/backup/zarafa/

[* snip *]

28-Nov 06:55 backup01_dir JobId 3325:       
/mnt/backup/zarafa/KH/KW40/KH.data.zbk
28-Nov 06:55 backup01_dir JobId 3325: Error: Bacula backup01_dir 3.0.3 
(18Oct09): 28-Nov-2009 06:55:30
  Build OS:               i686-pc-linux-gnu ubuntu 8.04
  JobId:                  3325
  Job:                    nas01_zarafa_verify.2009-11-28_06.00.00_50
  FileSet:                nas01_zarafa
  Verify Level:           VolumeToCatalog
  Client:                 backup01_fd
  Verify JobId:           3320
  Verify Job:             nas01_zarafa
  Start time:             28-Nov-2009 06:29:58
  End time:               28-Nov-2009 06:55:30
  Files Expected:         402
  Files Examined:         179
  Non-fatal FD errors:    0
  FD termination status:  Error
  SD termination status:  Error
  Termination:            *** Verify Error ***

28-Nov 06:55 backup01_dir JobId 3325: Begin pruning Jobs.
28-Nov 06:55 backup01_dir JobId 3325: No Jobs found to prune.
28-Nov 06:55 backup01_dir JobId 3325: Begin pruning Files.
28-Nov 06:55 backup01_dir JobId 3325: No Files found to prune.
28-Nov 06:55 backup01_dir JobId 3325: End auto prune.


All full Backups on the same tapes a verified OK. So I guess bacula expects to 
find something else on the tape due to the virtual full...


Kind regards

Julian

-----Ursprüngliche Nachricht-----
Von: Fahrer, Julian [mailto:julian AT fahrer DOT net] 
Gesendet: Mittwoch, 18. November 2009 11:47
An: bacula-users AT lists.sourceforge DOT net
Betreff: Re: [Bacula-users] virtuafull and verify

What would be ok, since the idea behind the verify job is to verify the last 
job...
Running a verify before the virtualfull would check the original job.

I just compiled it with the sql statement below. The verify just selected the 
right jobid...

-----Ursprüngliche Nachricht-----
Von: Graham Keeling [mailto:graham AT equiinet DOT com] 
Gesendet: Mittwoch, 18. November 2009 11:42
An: bacula-users AT lists.sourceforge DOT net
Betreff: Re: [Bacula-users] virtuafull and verify

On Wed, Nov 18, 2009 at 11:29:38AM +0100, Fahrer, Julian wrote:
> Hey,
> 
> yes it seems tob e the same problem.
> The sql statement in the sub "db_find_last_jobid" is:
> "SELECT JobId FROM Job WHERE Type='B' AND JobStatus IN ('T','W') AND "
> "Name='%s' ORDER BY StartTime DESC LIMIT 1", Name);
> 
> I think changing it to something like
> 
> "SELECT JobId FROM Job WHERE Type='B' AND JobStatus IN ('T','W') AND "
> "Name='%s' ORDER BY StartTime DESC, JobId DESC LIMIT 1", Name);
> 
> would fix the problem for me right now. But I am not sure what problems would 
> occure then...

The obvious problem is that you wouldn't be able to verify the original job. :)

> 
> Thanks for submitting my mail to the devil list


------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
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>