I get that's how it supposed to work -- and what I expected -- but 
that's not at all what was happening. Here's the behavior I saw:

- Tape magazine #1, loaded with 12 tapes labelled 000001L5 - 000012L5, 
was in the tape library; 11 marked as Full, 1 marked as Append
- I followed the documentation to swap the magazine out (unmount, change 
magazines, update slots, mount)
- Now Tape magazine #2, loaded with 12 more tapes labelled 000013L5 - 
000024L5, was in the tape library; all 12 marked as Append
- All tapes from magazine #1 were marked InChanger=0; all tapes from 
magazine #2 were marked InChanger=1
- Spool jobs would run, and Bacula would call for the one tape from 
magazine #1 (000003L5) that was marked as Append, despite having a full 
magazine of tapes in the library ready to go.
- I could remount, and specify the slot to use, and it would pull a tape 
from magazine #2, but as soon as that tape was full, it would ask for 
that same tape from magazine #1 again.
- Setting Enabled=1 for tapes in the library and Enabled=0 for tapes not 
in the library invokes the expected behavior

So in this case, there were no volumes that needed to be recycled; all 
were new, labelled, defined in the pool, and marked as Append.

Thoughts?

Kevin B. Zimmerman

On 07/03/2013 01:56 PM, John Drescher wrote:
> On Wed, Jul 3, 2013 at 1:53 PM, John Drescher <drescherjm AT gmail DOT com> 
> wrote:
>>> http://www.bacula.org/en/dev-manual/main/main/Autochanger_Resource.html#SECTION002950000000000000000
>>> "...Bacula will as a preference, use Volumes that it knows to be in the
>>> autochanger before accessing Volumes that are not in the autochanger.
>>> This prevents unneeded operator intervention."
>>>
>>> Behavior doesn't quite match docs in this case, as it appears the
>>> InChanger flag is ignored in favor of the Enabled flag when deciding on
>>> which tape to request next, at least as far as Spool jobs are concerned
>>> (that's all I've really been using so far).
>>>
>>>
>>> So, bug report?  Doc fix?  Both?
>>>
>> I say Doc addition. Since I do not believe the Doc is wrong. Just that
>> the information you need is in 2 different places.
> What I mean by this is bacula does honor the InChanger flag if there
> is an appendable volume in the changer. It however will not recycle a
> volume in the changer if there is a volume outside the changer that is
> appendable.
>
> John


------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users