does my servers require an upgrade?

[kagie]

Hello,

Could you pls have a look at 3 day statistics:

http://img33.imageshack.us/img33/9901/servdb.png

http://img42.imageshack.us/img42/2294/servap.png

and please let me know your opinion about the system utilization.

1) is the 'disk busy' graph telling that i/o waits is the problem in the system or not yet?

2) is memory enough? is the "pagspace paging" graph telling that memory amount is still enough?

CPU don't seem to be haevy loaded.

Systems have 12GB RAM, 4 x P5@1.5GHz, system on local drives and datavg on SAN HDS storage.

servap uses NFS filesystems served by servdb (both servers are members of HACMP cluster).


thank you in nadvance, Kris

 

[psi07004]

Rule of thumb for performance...
Good is the ideal value
Bad is where performance is affected
Very Bad is where the performance is serverly affected and
the component becomes a bottleneck.

Processor...
Good, 75% or less busy
Bad, 85% busy
Very Bad, 90% or more busy

Disk...
Good, 30% or less busy
Bad, 40% busy
Very Bad, 50% or more busy

Memory...
Good, zero paging
Bad, 20 pages/sec/proc (large systems can do more)
Very bad, >20 pages/sec/proc

Network...
Good, 30% or less busy (minimizes collisions)
Bad, 30% to 40% busy
Very bad, 40% or more busy

From a quick look at your graphs it seems you are ok.

Are you having a particular performance problem?

 

[mrn]

Paging is high and so is NFS usage.

The disk i/o is on hdisk 0 and 1 so more than likely the paging is causing this.

Get a trail of sarcheck @ sarcheck.com should tell you if it's memory bound or if it's leaky code that's causing your problems.

Mike

"Whenever I dwell for any length of time on my own shortcomings, they gradually begin to seem mild, harmless, rather engaging little things, not at all like the staring defects in other people's characters."