This is to address the recent notes about the dual gripper in the 3494, and
adsm's not utilizing it.  We are currently testing the next adsm ptf which has
change to our dismount/mount sequence that *should* make use of the dual
gripper and increase the throughput for mounts (especially if the 3494 is
partioned with another host/application).  The change is that ADSM will not
wait for dismounts to complete before issuing a mount (it WILL wait for the
rewind-unload from the drive so that there will not be a sequence check).  This
is what's required to queue up a mount on a drive that is still loaded by the
last transaction.  It also increases the priority of the dismount, which is
normally low.  In a partioned library, if the other application is already
doing dismount/mounts this way, ADSM will now get a more balanced response to
its mount requests.

Unfortunately, this change will not cause the LM to  utilize the dual gripper
when inserting cartridges from the mailbox E/E port.

Craig Bell
ADSM Server Development