=> On Tue, 27 Feb 2001 10:44:04 +1100, Carl Makin <carl AT XENA.IPAUSTRALIA.GOV 
DOT AU> said:

> I do think our D40 drawer (5 x 36gb disks) was setup poorly.  I've been over
> the hardware config with our IBM engineer and he thinks it's probably not
> running at the full loop speed (SSA160).  We have a mixture of 40's and 160s
> on the same loop which is crossed over in a HA configuration between two
> serialraid 160 adaptors.  Fast-Write is disabled, I just checked.

Well, then that's quite a few factors that will give you sub-par SSA
performance.  I'd like to address one factor I haven't seen hit on yet: your
RAID setup:


This is of course affected by your environment, but I would reccomend that you
_not_ raid your disk pools: Most times, the disk pools are only holding unique
copies of data for a very brief period: From the beginning of a backup until
the end of a copy.  This is not a very high exposure.

If you split out your RAID, you may not only increase your availiable space by
33% (or 25% if you have ommitted a hot spare), but you can make sure each disk
volume occupies only one spindle.  (not necessarily one volume per spindle:
36G volumes may be too big for your taste)

If you do that, you can actually watch (on e.g. monitor) TSM stripe disk
accesses between the spindles.  It's heartwarming.

Further, Most SSA adapters have a pair of loops: I'd reccomend you split your
storage loop from your DB/Log loop.



I would be ...very surprised... if you found that a well-configured SSA
install did anything but leave a Shark twitching on the sand.



Allen S. Rout
NERDC TSM type.