We have had very similar performance woes since our upgrade to 5.0 from 4.0
less than 2 weeks ago.  We will be upgrading to a 995-200 from a 992-200 in a
few weeks, but we asked HP to help us identify the problems now, before the
upgrade.
They have come up with the following 2 possible solutions:
  1.  We changed our "queues" so that the job "queue" and the session "queue"
did not overlap because of some known problem with sessions that wait on
message files.  Evidently if a session drops down to the bottom of its
"queue," it doesn't get bumped back up to the top of its queue like 4.0 did.
Therefore it is now lower if priority than some jobs if your 2 queue's
overlap.  That is the explanation that we received from HP, don't ask me to
explain why it would make a difference on our system, but we did make the
change and things might be a little better.
 
  2.  This is the big one.  We are set to receive a patch this week which is
suppose to fix a known 5.0 problem which causes performance problems when
doing a store or a restore from a tape.  This problem we can feel.  I can tell
from my terminal when a store is being done in the computer room because of
the awful response time.  Until we receive the patch, we do all stores and
restores after hours.
 
This is the little bit of knowledge that we have gained from HP since the
upgrade, if anyone has any further info please respond.
 
Clint Schwartz
Programming Manager
Memorial Health System
South Bend, Indiana
[log in to unmask]