[Orca-users] RE: Orca-users Digest, Vol 6, Issue 15

[Liston Bias]

On Thu, 19 Jun 2003, JV wrote:

> From: "Garrett, Matt M SITI-ITDIEEE" <matt.garrett at shell.com>
> Subject: [Orca-users] Orca using over 40% CPU on Server
>
> >However running on a Sun Solaris 8 system , Ultra 30 , UltraSPARC-II 296MHz
> >CPU , 512Mb memory
> >The Orca process seems to all ways use over 40% of the CPU
> >The only reason it does not use 100% is other monitoring process (Nagios)
> >need to run as well.
>
> >At the moment I am only monitoring 53 clients.
> >This is about to be stepped up to 300 clients.
>
> >Any suggestions or idea's would be very helpful
>
>
> Matt and Liston,
>
> my 2 cents -
>
> my setup is 1 e250 with 2 400 Mhz cpus and 1.5 GB of RAM, with about 40
> subscribing orca clients. I found the bug with the duplicate disk graphs to
> be so very annoying and reduced my credibility in presentations to others,
> so I disabled them by commenting them out.
>
> Wallah! my "orca cycle" time when from 2.5 hours to 25 minutes. Since I am
> the only one who really looks at orca, I found this to be acceptable.
>
> I noticed with prstat -a that orca only ever uses 1 cpu, at 50%
> utiltization. Liston, how do you get 90% util on your v480R?

The individual disk stats don't have much value for us, so I dropped them
and it help tremendously as it did with you.

I run about 10 threads of orcallator for solaris and linux graph
generation.  I also do much activity in /tmp.

- Liston