Daniel Stenberg wrote:

> On Wed, 11 Dec 2002, Ralph Mitchell wrote:
>
> > Thanks, Daniel. What I really need is some way to measure how badly loaded
> > a web server is, and I think what I'm calling "page render time" may be
> > about as good as I can get.
>
> If I were to measure this I think I'd use a machine network-wise very close
> to the one I'd test (just to be able to rule out network delays) and then I'd
> count the full transfer time (- pretransfer) as a measurement, as the whole
> deliver period will be affected by the server load.

If only I could choose where my test box lives... The servers I'm hitting might
be in the same underground bunker, or they might be 250 miles away in Dallas.

So far I've been collecting total_time, starttransfer and pretransfer, and
stuffing them in to a rrdtool database so that I can generate graphs to get an
idea about what's happening. Worst spike so far was a 2.3 second starttransfer
with a corresponding 0.3 second spike in pretransfer. It'll be interesting to
compare the times when we get a "my server is struggling" call.

Ralph

-------------------------------------------------------
This sf.net email is sponsored by:
With Great Power, Comes Great Responsibility
Learn to use your power at OSDN's High Performance Computing Channel
http://hpc.devchannel.org/
Received on 2002-12-11