Another possibility might be the Switch Controller; the logic is then in how the value is calculated. One of the switch entries can be a dummy sample, e.g. Test Action with no wait; this allows the Switch Controller to do nothing.
== For a basic server load test, it should be sufficient to just ensure that the overall load is representative. The effect of individual random behaviour on the overall load tends to get less and less as more sessions are used. On 26/05/2009, Noel O'Brien <[email protected]> wrote: > Yes, but won't that execute the same requests each iteration (just like the > proxy)? What I was trying to achieve was to simulate the randomness of > choosing different paths through the client application, which leads to > realistic groups of API calls being made to the server > > Either way I still had to re-organise the requests into Random/Throughput > Controllers to simulating random branching with weighting. > > Regards, > Noel > > > ----- Original Message ----- > From: "Peter Lin" <[email protected]> > To: "JMeter Users List" <[email protected]> > > Sent: Tuesday, 26 May, 2009 17:12:37 GMT +00:00 GMT Britain, Ireland, Portugal > Subject: Re: StackOverflowError > > > If you want to simulate "real production traffic", one option is to > use the access log sampler. > > I wrote that sampler so that I could simulate production traffic. It > might be easier than using a test plan that generates random requests. > > peter > > On Tue, May 26, 2009 at 12:08 PM, Noel O'Brien <[email protected]> wrote: > > Done: > > https://issues.apache.org/bugzilla/show_bug.cgi?id=47269 > > > > Regards, > > Noel > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > > > > > -- > Regards, > > Noel > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
