Hi
We used to have some big IBM servers with AIX running on int (vers. 4.3? cannot remember anymore).
Well, it looks like your CPUs are not really 100% busy - there are just "wait"s...
Quote:
Wait 92.0 |##########################
|
... (this is reported ), so the CPUs are just waiting that stuff gets read or written to/from HDDs but are still available for real processing, and the two HDDs show that there is actually activity there going on:
Quote:
Disk Busy% KBPS TPS KB-Read KB-Writ Steals 327 % Comp 86.
hdisk0 100.0 1308.2 298.5 206.0 1102.2 PgspIn 95 % Noncomp 14.
hdisk1 100.0 1278.2 294.5 180.0 1098.2 PgspOut 276 % Client 14.
|
So, you just have to find out which process is fiddling around with the HDDs.
Btw, do you use
nmon to monitor the machine? I used it very often when I used to monitor our servers, and it was really great to have a single utility which shows everything. I am still using it now in Linux, and I am really missing it with our new Solaris-servers.