Clever Geek Handbook

Testing load balancing with Jmeter

I have successfully set up a test plan in Jmeter that can execute 500 threads and make multiple HTTP requests on the same machine and integrate with Hudson. So let me say that I want to increase my threads to 6000.

To begin with, I created several machines to accommodate a smaller total number of threads, for example 1500. I use Jmeter's built-in run function in distribution mode. I enter 1500 threads into the test master plan. So here are my $ 10 questions:

  • Why does each client start 1500 threads and not intelligently share the total number of threads with the controller / master? (Have I configured my test plan incorrectly?)
  • Is it possible that Jmeter also keeps track of the number of machines allocated to them and also keeps track of the sequential count of the counters?

I completely misunderstood the concept of executing a load in propagation mode (I think so)? Do I need to manually set up variables, threads, loops, etc. And run them separately and manually on each machine if I want to split the load?

Thanks for reading!

+3


source to share


1 answer


  • This is not how it works, distributed testing has a copy of the test plan and sends it to the jmeter-servers, so if you want a total of 1500, set the number in the plan to 1500 divided by the jmeter-server number
  • Each jmeter server keeps track of its own number, but does not report it to the controller. But you get this in the output report

Read this:



With JMeter, you usually need to replicate in this mode:

  • data used by CSV

  • jmeter.properties and user.properties

But for the test plan, you only need to have it on the controller.

+3


source






More articles:


All Articles