NBU 3.4.1, HP-UX 11i masters in a 3-way Veritas ClusterManager cluster, the
master is also currently the media server, that may soon change. Library is
an ADIC Scalar 10k currently has 10 - IBM magstar LTO(1) drives, we are
adding 10 additional drives, but these are LTO-2.  Because of the forward
incompatibility we will have with the LTO2 tapes possibly being used on an
LTO(1) drive, which will fail and cause frozen tape snarls, it looks like we
need to set up another storage unit for the new drives, as well as another
volume pool for the LTO2 tapes. My question is: can anyone describe the
"locking" mechanism in NBU by which simultaneous requests to the same robot
which minds two different storage units are managed?  Is this a "gotcha"? Is
this a "don't do that"?  Will we just get a bunch of 219 storage unit
unavailable errors or is this supported?  Has anyone else attempted to do
this?  Thanks all!