Subject: New AIX5.3 ML3 Centralized Performance Monitoring Tool

Audience: All

Date: October 10, 2005

The new "topas -C" command (AIX5.3 ML3) provides a centrailzed performance monitor for all partitions in a Power5 server.

The output looks like:

Topas CEC Monitor             Interval:  10             Mon Oct 10 07:54:15 2005
Partitions     Memory (GB)               Processors
Shr:  4        Mon: 2.3  InUse: 2.0      Shr:1.4  PSz:  2   Shr_PhysB:  0.02
Ded:  0        Avl:   -                  Ded:  0  APP:  2.0 Ded_PhysB:  0.00

Host         OS  M Mem InU Lp  Us Sy Wa Id  PhysB  Ent  %EntC Vcsw PhI
-------------------------------------shared-------------------------------------
p520-sr2     A53 S 1.0 1.0  4   0  1  0 98   0.00  0.20   2.3  393   0 222
p520-sr5     A53 S 0.4 0.4  4   0  0  0 99   0.00  0.20   2.0  367   0 093
p520-sr4     A53 S 0.4 0.3  4   0  0  0 99   0.00  0.20   1.9  297   0 479
p520-sr3     A53 S 0.5 0.3  4   0  0  0 99   0.00  0.80   0.5  299   0 950




------------------------------------dedicated-----------------------------------

The %EntC column shows the utilization rate of each partition as a percentage of its entitled capacity.

Comment: I initially had problems with the new command (hangs). It was fixed by changing the following line in /etc/inetd.conf from

xmquery dgram udp wait root /usr/bin/xmservd xmservd -p3

to

xmquery dgram udp wait root /usr/bin/xmtopas xmtopas -p3

Refresh the daemons by running: "refresh -s inetd", and "kill -HUP xmservd".