You can mix any thing with anything....except you can't mix ficon and scsi (FCP) on the same LCU. (You can physically LPAR a DS6800 and you can logically LPAR a DS8000).
When you add volumes to a LCU, you specify the device number (IOCP translates device numbers to our 390 addresses). I would test adding one device and check the address produced. I always get it confused as on one panel, the number entered is in decimal, and on the other panel, the number entered is in hex. Sometimes it takes a while to find the drive I just created. You are not going to damage any existing volumes. Except when you start deleting the wrong addresses <G>. Tom Duerbusch THD Consulting Law of Cat Acceleration A cat will accelerate at a constant rate, until he gets good and ready to stop. >>> Dave Reinken <[EMAIL PROTECTED]> 6/6/2008 12:38 PM >>> I've got a 2105-800 coming in to replace a 9393 and 2105-F20. The 9393 needs to be pulled to make room for the 2105-800, so I am squishing what I need from it onto the 2105-F20 for now. The F20 is currently defined with 8 LCU's of 40 3390-9s each, leaving an extra 88,456 cylinders in each LCU. I need some 3390-3s. I'd like to stick them onto the end of LCU7 so that it doesn't trash my addressing, but there is stuff in LCU7 that I need. 1) I can mix 3390-9 and 3390-3 within an LCU, right? 2) When I add volumes to an existing LCU (say 26 3390-3s to the existing 40 3390-9s in LCU7) is it going to nuke the existing 3390-9s? Unfortunately, the data that I _don't_ need is in LCU0-4. I suppose I can deal with it and just readdress it in the IOCP by LCU to make it as painless as possible, but it sure would be nice if I could just tack 26 3390-3s onto the end for a couple days until the 2105-800 is in place.
