We are in the process of performing stress test and we noticed that the
IO wait time has suddenly increased by a very high percentage. There
are 6 processors and 24 disks. Initially the wait times were very
negligible, however off late it's gone upto 70-80%.
Wld appreciate if any of you who have encountered such a situation
could throw some light on the possible reasons for this. Do update if
any more information is required.

read the perf docs .. wait i/o means the cpu is idle with at least 1
i/o outstanding for that cpu .. high wait i/o CAN indicate disk
subsystem issues, but the real test is to benchmark your disk subsystem
and test i/o throughput and response time. wait i/o is "more often than
not" a red herring.

Rgds
Mark Taylor

what's the best way to figure out the problems with disk IO. do we have
a utility for this!!!!!

Sure.

You can use filemon to identify top I/O activities:
filemon -O lf -o filemon.out # start capture
trcstop # stop capture

Use -T option if you get "buffer overflow" messages

0xbeef