[aklug] Re: Reasonable average server load?

On Wed, 23 Jun 2010, Christopher Howard wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Question for you sys-admin types: If you were running a farm of web
> servers, what would you think would be a reasonable average CPU load?
>
> Reason I ask: I manage my web sites with Bluehost over ssh, but more and
> more lately, when I log in over SSH, everything I do on the connection
> runs really slow. For example, starting Emacs takes so long that I
> finally gave up and switched to ed, and even ed needs one to two seconds
> per command.
>
> Anyway, I want to officially complain about it, but I also want to look
> like I know what I'm talking about. I can run top, and it shows a whole
> lot of PHP and MySQL processes, and a CPU load oscillating between about
> 20% and 50%. Is that an ideal CPU load? Is there better statistic that I
> should be looking at?

System load is a good basic indicator and should ideally not get over 1.0,
since that means you have processes in the run queue that are ready to run,
but are blocking.

If you want to isolate what the specific problem on the box is you should be
using vmstat(8) and iostat(8). Between those two you should be able to see
if the problem is due to a bottleneck in either the hardware architecture or
an artifact of the software running on it. In a more perfect world you'd be
running SAR reports. If you have access to that you'll have all that data
and more.

         --Arthur Corliss
           Live Free or Die
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Wed Jun 23 11:25:02 2010