Re: [Bacula-users] Fatal error: askdir.c:355 NULL Volume name. This shouldn't happen!!!

[Kern S. <ke...@si...>]

On 07/22/2014 11:34 PM, Clark, Patricia A. wrote:
> Sadly, this error is still present in v7.0.4.  There is another error, too, where there is no writeable media in the tape drive.
> Fatal error: spool.c:263 Fatal append error on device "LTO5-2" (/dev/tape-admin2): ERR=tape_dev.c:161 Unable to open device "LTO5-2" (/dev/tape-admin2): ERR=No medium found
> Both of these are a result of race conditions occurring on tape libraries with multiple tape drives, one job per drive, and several backup jobs running nearly simultaneously.  Frequently, the storage daemon requires a restart in order to release the tape drive affected by the error since the SD still sees a running job, however, the director has already indicated the error, sent the notification, and does not have a running job to cancel.  On a busy system, restarting the SD can be problematic, and losing one or more tape drives is also a problem.
>
> My question, why isn't Bacula happy with a writeable volume from the same pool if one is mounted in the tape drive assigned to a job?

I don't know the answer to the question otherwise I would fix the
problem. It could be for a number of reasons one being that the Volume
has been pruned or otherwise expired.  Most often triggering the problem
has something to do with your particular configuration, which judging by
where you work, is probably large and/or complicated.

If you can provide more information on this, it is worth posting a bug
report.  I will need copies of you bacula-dir.conf and bacula-sd.conf
(with passwords removed please) as well as a "status storage", "status
dir", "list Volumes", and "llist volume=xxx" for each Volume that is
"stuck" on the SD when the problem occurs.  This will show me a bit what
is going on.  Given the complexity of the problem, you will probably
need to install gdb on your SD so we can examine the internal state of
the SD when it gets into this condition.

Are you doing Volume pruning?  If so, is your Volume Retention Period
set to an exact multiple of 24 hours.  If that is the case, set your
Volume retention to a multiple of 24 hours minus 2 hours, then sure that
you properly reload your configuration and that you update all your
current Volumes.  If the first two conditions (pruning) are true, this
has about 95% chance of eliminating your problems. If this eliminates
your problems, I won't need a bug report.  For the case I have just
described, I know the problem, but I am not yet 100% sure what the best
way to fix it is.

Best regards,
Kern
>
> Patti Clark
> Linux System Administrator
> R&D Systems Support Oak Ridge National Laboratory
>
>
> ------------------------------------------------------------------------------
> Want fast and easy access to all the code in your enterprise? Index and
> search up to 200,000 lines of code with a free copy of Black Duck
> Code Sight - the same software that powers the world's largest code
> search on Ohloh, the Black Duck Open Hub! Try it now.
> http://p.sf.net/sfu/bds
> _______________________________________________
> Bacula-users mailing list
> Bac...@li...
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>