Hello TSM'ers

I have a strange problem, or is it 'Working As Designed?'


Using a manual SCSI library with 9 slots, there are 4 tapes checked in as
scratch defined to stgpool dailypool, 4 tapes chacked in as scratch defined
to stgpool weeklypool, 1 tape checked in as scratch defined to no storage
pools.


When issuing a backup db command I expected TSM to move straight to the
undefined volume and use it as it's scratch volume for the DB backup.


What we are seeing however, is the DB backup process wait upto 18 minutes
for a scratch volume as TSM mounts each volume in the library in turn, tries
to use this as the scratch volume for the DB backup but fails with the
error;

25/10/01   14:42:25  ANR8337I DLT volume ABD008 mounted in drive DRIVE1
                      (/dev/rmt/2mt).
25/10/01   14:42:25  ANR8448E Scratch volume ABD008 from library SUNLIB
                      rejected - volume name is already in use.


and so on untill it finds the one free undefined scratch volume in the
library and backs up to that volume.

Now, why does TSM do this when it knows the status of the various tapes in
the various slots in the library?


Help Much Appreciated !

Matt.

_________________________________________________
Matthew Warren. GZPTXW.
UK Technical Support; Storage Administration Group.
Wavendon Tower, Milton Keynes.
Office / +44(0)1908284481
Mobile / +44(0)7790492034
__________________________________________________