Hello all,
     We backup an ORACLE ERP that is about 350GB.  We back it up by
running a script that uses  16 concurrent sessions, each one backing up
a series of File Systems, all sesions going to disk pool.  It is
compressed by the client.    This method works well.  We backup the data
in less than 2 hrs and it looks like 130GB of compressed data.   I know
the data will be single threaded through 1 tape drive to tape during
migration, because it is all coming from the same server.  The offsite
copy is then made running tapepool to copypool, (yes I know it would be
better to make the offsite from disk first but I do not have enough disk
for enough time to do that).    I know that our 9840 tapes will be
getting about 20GB each so there should be 5 or 6 tapes (at least)
containing this servers backup.
    What would be the optimal way of doing the RESTORE of this data?
Ideally I ask TSM to just use 6 drives and all is well or know where the
break points are on the tapes.  Has anyone worked (and tested) this out
yet?  Our current attempts take about 18-24 hours 
Thanks in advance
Matt