[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bacula-devel] Multiple fd plugins in one fileset: Restore problems

Hello Bastian,

Since this is moderately complicated, to help me analyse this, could you 
collect everything together into one email:

1. Your bacula-dir.conf file or the critical pieces
2. The Job listing of the backup output.
3. The exact command you used for the restore
4. The job listing of the restore output
5. An "ls -l" of the original files
6. An "ls -l" of the restored files
7. The -d50 of the restore.

If you want, you can send it by email or create a bug report.  The advantage 
for me of a bug report is that everything is always together -- with email, 
it is too much work to search through what came in a few days ago, and 
threads don't always help.

With all that in hand, hopefully I can either find the problem or duplicate 

Best regards,


On Tuesday 30 September 2008 14:31:12 Bastian Friedrich wrote:
> Hello Kern,
> On Tuesday 30 September 2008, Kern Sibbald wrote:
> > Something is going on that I don't quite understand.  Perhaps you are not
> > asking to restore all the files, and the code is going into the plugin
> > code when I believe it logically should not.
> Sorry that I did not myself clear.
> Yes, I was - in some of the cases described - not restoring all files, but
> only a subset.
> The enviroment described in my original post was not fictional; I use the
> "in" and "out" scripts described there to reproduce the problem.
> > Anyway, I have just committed another patch to the SVN that will probably
> > fix the problem.
> Thx a lot! The segfault is gone now, so the problem you probably were
> mainly addressing is solved.
> I do still have problems restoring data contained in multiple plugin
> streams, nonetheless, such as described in cases 1 thru 4 of the original
> post.
> > I've added an extra line of debug code that should help
> > me see what is really happening, so even if the bug is now fixed, I would
> > appreciate it if you would send me the output from a FD -d50  run.
> Attached for case 4 (no backtrace this time :))
> Side note: the 28 bytes reported by restore.c to be restored would be the
> correct number ('hello world X<CR>' is 14 chars). Unfortunately, all but
> the last "consumer" gets an empty input. Tested with my "real life" case,
> and the described testcase.
> Thx a lot once again for your efforts!
>    Bastian

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
Bacula-devel mailing list

This mailing list archive is a service of Copilotco.