Another vote for ab. I'd drop your provider a line in advance and let
them know out of courtesy, and to cover your ass on the offchance it
goes legal.

Bear in mind if you're running on a shared host your results are going
to be dependent on other things like how much machine load the host is
experiencing, and depending on how their network provision there might
also be an effect on latency/throughput. (SmokePing is useful for
tracking this, http://oss.oetiker.ch/smokeping/ You might even watch
for an effect during your tests.)

Don't forget to profile your app at some point :-)

Paul

On Thu, May 15, 2008 at 12:01 PM, Lyle <[EMAIL PROTECTED]> wrote:
> Hi All,
>  I'm nearing the end (finally) of a huge update (more of a rewrite) a
> software project (over 10,000+ lines).
>
> It's CGI based, tracking clicks. Now I want to run some proper
> benchmarks so I know exactly how much traffic the software can handle in
> clicks per second...
>
> I'd like a report of the number of clicks sent per second over say, an
> hour period and what the server load was for each minute.
>
> I'm assuming that I need 2 machines, with one send clicks to the other.
> I could use my US server to call my UK server, but I don't know what
> software to use and my searches haven't brought up much :-(
>
> Also I concerned that this is similar to a DOS attack? Is there a change
> my hosting providers will cut my connection or block my IP once I start
> the test? I could setup a couple of local machines, but I'm guessing the
> tests wouldn't be that accurate as they are conducted over a LAN rather
> than the net.
>
>
> Any and all advice and pointers welcome.
>
>
> Lyle
> _______________________________________________
> BristolBathPM mailing list
> [email protected]
> http://mailman.bristolbath.org/mailman/listinfo/bristolbathpm
>
_______________________________________________
BristolBathPM mailing list
[email protected]
http://mailman.bristolbath.org/mailman/listinfo/bristolbathpm

Reply via email to