(TSM server 5.1.6.5 on AIX 5.1 with essentially standard options)

>From observing DSMSERV's I/O behaviour via "topas",
I get the impression that it *writes* to the database volumes
(currently 14 files each covering a whole JFS-mirrored SSA disk)
in essentially single-threaded fashion, i.e. never exceeding
the I/O rate that you would expect from a single disk.

This is quite different from the *read* behaviour, where
many disks (= DB volumes) are being used in parallel.

Is there some tuning option that would allow for a higher
write I/O rate, or is the sequential processing of writes
fundamental to TSM's database operation?

I didn't find anything pertinent in the "Quickfacts" ...

What write I/O rates do you see?

Wolfgang J. Moeller, Tel. +49 551 201-1516/-1510, moeller AT gwdvms.dnet.gwdg 
DOT de
GWDG, D-37077 Goettingen, F.R.Germany     |    Disclaimer: No claim intended!
http://www.gwdg.de/~moeller/ ---- <moeller AT gwdg DOT de> ---- <w.moeller AT 
ieee DOT org>