Hot Diggety! Kauffman, Tom was rumored to have written:
> You should be able to do this -- I'd suggest shutting down TSM for an hour
> or so and play with the 3584 front-panel menu for configuration. I question
> the approach, however.

I later decided to just keep it as one logical library in order to allow
all stgpools fetch tapes from the same scratch volume pool on an as needed
basis, per another useful recommendation. In fact, I got a good number of
useful feedback, each addressing a different aspect, so I later incorporated
them all into the eventual design for the new setup.

It's coming along rather well now... but no shortage of problems just yet ;)
Today's new one: rmt8 (one of the 3584 drives) looks completely fubar'd -
not responding to VPD inquiries or SCSI inquiries or anything. Lovely. And
it's brand new! Then trying to load AIX 5.1 on the new server is the stuff
that nightmare factory puts out. ;) But I'll eventually get everything perfect.
(the 6H1 machine has a CDROM firmware bug that prevents a CD installation;
NIM install trips over a nasty bug that prevents it from getting far, etc.)

Past week, I've been having great fun with the 3575's SMC board knocking itself
offline -- and this is a replacement board, no less. (3rd, I think?) Can't
wait to get this new setup finished!

> Realisticly, if all that get checked out are copypool tapes and TSM database
> backups there shouldn't be a problem. I control that by naming conventions.
> I have six (I think) copy pools, all of which end in -COPY; my checkout
> process takes the result of 'q vol stg=*copy acce=reado,readw', checks them
> out, and changes the access to offsite.

Sounds like a winner. Thanks ;)

-Dan