actually my benchmarks against tomcat for static files was using
Httpsampler2, which uses apache Httpclient.  testing a blank page
really only tests the parsing of HTTP header and the OS socket
performance.

In practice, on real webpages, JMeter will do just fine.

On 6/17/05, sebb <[EMAIL PROTECTED]> wrote:
> Might be worth trying the HTTPClient HTTP Sampler, which uses the
> Apache HTTPClient implementation.
> 
> Also, running non-GUI is likely to help, as far fewer resources are needed.
> 
> S.
> On 6/17/05, Peter Lin <[EMAIL PROTECTED]> wrote:
> > if you look at the static file benchmark I ran several months back,
> > you'll see that JMeter does have limits and doesn't handle files under
> > 1K effectively.  I haven't had time to track down the cause. For files
> > larger than 1K, JMeter matches the performance of apache AB. you can
> > see the raw results here
> >
> > http://cvs.apache.org/~woolfel/tc_results.html
> >
> > the article is on Tomcat's resource page.
> >
> > peter lin
> >
> >
> > On 6/17/05, Ramazan Yildirim <[EMAIL PROTECTED]> wrote:
> > > We are using JMeter 2.0.3. There seems to be a problem with the Http
> > > Sampler component. The test was pretty simple, and we did the same
> > > experiment with a tool that we have written using Apache http client.
> > > What we did is we have written an empty jsp page and called it using
> > > JMeter and our http client tool. We have used 40 threads and used only
> > > one listener on the JMeter test.  The throughput was around 1600/min.
> > > The same test with the http client tool has yielded 10000/min.
> > >
> > > The client computer was a Celeron 1.7Ghz, 1GB memory. And JVM memory
> > > parameters were -Xms256m -Xmx256m.
> > >
> > > What happens on the JMeter side is weird interruptions when calling the
> > > web server.
> > >
> > > I think I have tested it using no listeners, the same interruptions were
> > > still there.
> > >
> > > I never have tested it with a lower thread size.
> > >
> > > What's might be wrong with this setup? Or is this the expected outcome?
> > >
> > > Regards,
> > >
> > > Ramazan YILDIRIM
> > >
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > > For additional commands, e-mail: [EMAIL PROTECTED]
> > >
> > >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to