Forum Discussion

anuthakur's avatar
anuthakur
Occasional Contributor
12 years ago

Problem controlling the number of concurrent requests

Hi

We are currently evaluating the trial version of LoadUI Pro for our project. This is what we need to achieve -

We have a webservice that we need to test for performance. For that we want to gradually increase(basically ramp up) the number of concurrent users and see what is the number of concurrent requests our service can handle at which the throughput becomes constant and doesn't grow further.
We need to analyze the max load that our service can handle without growing the response time.

We tried using SoapUI runner and Ramp generator to achieve this. But the ramp generator doesn't seem to control the concurrent requests that the SoapUI runner would fire. It does not ramp up the number of requests.

How do we achieve this.

Thanks
  • Hi!

    Please try the Ramp Sequence Generator. It ramps up the rate of created users.

    LoadUI mostly uses Rate of incoming users instead of Concurrent users in the system. We believe that this is a more modern and better way to load test. Read more about this here.

    Regards!

    Henrik Olsson
    LoadUI Product Owner
  • anuthakur's avatar
    anuthakur
    Occasional Contributor
    Thanks for the reply Henrik.

    I am newbie and not sure if I have correctly understood how LoadUI works. I have been referring the docs -
    http://www.loadui.org/Getting-results/s ... rence.html
    http://www.loadui.org/Runners/soapui-ru ... oncurrency


    The things I would like to understand is how the counters on the Runners display function.

    Say I have set a generator which generates load @ 10/sec. With this rate when and how the below counters on the Runner component will change -
    Requests
    Running
    Completed
    Queued

    How would I know when my server has reached the max load and can handle no more incoming requests. I need to identify the max concurrent requests which my server can handle without queuing the requests.

    Thanks
  • anuthakur's avatar
    anuthakur
    Occasional Contributor
    Hi,

    If I have understood Fixed Load generator properly, it would always run x requests concurrently. And when all x requests are complete it would trigger the next x concurrent requests. Is that correct?

    My requirement is to generate a load with a number of concurrent requests and then gradually increase the number of concurrent requests over a period of time to see what is the max concurrent requests my server can handle.

    With Fixed Load generator I could only trigger a fixed number of concurrent requests. Is it possible to somehow ramp up the number of concurrent requests with load ui.

    Thanks
    Anu
  • anuthakur's avatar
    anuthakur
    Occasional Contributor
    I think with Fixed Load generator its x threads running simultaneously. As soon as a thread completes, new thread is started to keep the running count always equal to the load set.