I have 4 3590 drives in a 3494 library.   ADSM clients are writing to disk pools
and data is migrated to tape.    Questions are:

1. Given that my disk storage pools are not large enough to hold all of the data
from clients without filling up a few times, how should I set my HIGH and LOW
Migration levels to get the best throughput ?

2.  Reading through the IBM 3590 operators guide, I see that they suggest
setting the tape block size (on AIX) to "0" (variable block size    What block
size are you setting for your 3590 drives ?

3.  If I change, the block size of my 3590 drives from "0" to some fixed value,
won't this cause problems for data backed-up with a variable block size ?