Go back to all articles

JMeter. Changing intensity with Constant Throughput Timer

Jul 26, 2020
2 min read

JMeter is a highly extensible, open-source platform that can help identify performance issues in your application. It is among the most popular tools for performance testing. You can read more about difference between PFLB, LoadRunner and JMeter.

So in JMeter, it is no longer possible to change the intensity, the number of threads and the duration of the test by standard methods after starting the test. However, it happens that the test is already running, but you urgently need to change a parameter (for example, I often forget to indicate the test duration), for which you need to stop the test, set database to the initial state, reboot the cluster, pull the original git commit and so on. Solution: configure and run BeanShell Server. With the help of BeanShell Server one can “play” with the following settings. Let’s illustrate it on the example of intensity change.

Variables

jmeter_loc – the full path to the root jmeter folder

host, port – jmeter’s host and port for the machine, where it is running

bsh_loc – folder, containing bsh-files

new_throughput – new intensity value, op/min

old_throughput – default intensity value, op/min

JMeter configuration and start from the command line

  • 01
    We create the file name.bsh: «setprop(“hits”,args[0]);»
  • 02
    In jmeter.properties search for: «BeanShell configuration» and comment out the lines «beanshell.server.port=9000» и «beanshell.server.file=../extras/startup.bsh».
  • 03
    After rebooting jmeter the server start line appears in the command line window:
  • 04
    Set target JMeter throughput to ${__P(hits, %old_throughput%)}.
  • 05
    If needed during the test we set the new throughput from the jmeter root folder in cmd to: «java –jar %jmeter_loc%\lib\bshclient.jar %host% %port% %bsh_loc%\name.bsh %new_throughput%».

Explanation

In the line «java –jar %jmeter_loc%\lib\bshclient.jar %host% %port% %bsh_loc%\name.bsh %new_throughput%» we call bshclient and indicate, which host and port are opened to the bsh-server, as well as which bsh-script needs a new parameter.

Example solution for a maximum search test

The project profile consisted of 20+ operation, so changing each intensity change manually is not fun. We wanted to find a maximally scalable and parametrized solution, which would allow us to use it for different projects with a different operation amount and from different loading machines.

Let’s illustrate it on the example of intensity change.

In the end we’ve written a “universal” .bat-file to change intensity,

jmeter example 1

where vals-line contains new intensity values, separated by letter “a”.

name.bsh:

jmeter example 2

In JMeter load test scenario we add a separate thread group and in it – jsr223-sampler. Constant timer is added to the thread group and the intensity change is fired by it. Base intensity values for each iteration are saved in User Defined Variables.

jmeter example user defined param

Function Helper explains, how to use intensity values syntax in Target Throughput.

jmeter example function helper

After the .bat-file has been executed, the command line stays open, so we see that the intensity values have successfully been changed.

jmeter example command

And that the command has looked like that:

java -jar C:/Users/User/Desktop/apache-jmeter-5.1.1\lib\bshclient.jar localhost 9000 C:/Users/User/Desktop/apache-jmeter-5.1.1/bsh\name.bsh 1344 2016 2016 240

We observe the intensity change in Transaction per Second:

jmeter example changing intensity

The intensity has been doubled.

Have a Project in Mind?​
We have been working on performance testing projects since 2008.
Drop us a line to find out what our team can do for you.
Get a quote You’ll hear back from our tech account manager in one day if not sooner

Conclusion

Using the described approach, you can increase and reduce intensity, thread amount and test duration, which saves time and helps to choose the optimal load options.

Table of contents

    Related insights in blog articles

    Explore what we’ve learned from these experiences
    5 min read

    6 AI Tools Performance Testers Already Use Today

    ai tools for performance testing preview
    Sep 17, 2025

    Performance testing tools already rely on AI methods. Monitoring platforms use anomaly detection, cloud providers forecast demand for autoscaling, and log systems group patterns with machine learning. These features are often treated as standard, but they are all examples of AI applied in practice. This article looks at six areas where testers already work with […]

    4 min read

    Top 7 Application Performance Monitoring Tools in 2025: Detailed Comparison

    best apm tools preview
    Sep 15, 2025

    Application performance monitoring (APM) tools have been created to control and handle the operation of different software solutions. They gather company-wide information, track important metrics — such as system response times, availability, and user interactions — and provide immediate insights into application inefficiencies. The business benefits of these instruments include quicker and more efficient operations, […]

    15 min read

    Top 10 Load Testing Tools for 2025: The Deep Dive

    best load testing tools preview
    Sep 9, 2025

    Effective testing is all about choosing the best load testing tool, which is just right for you. Top performance testing tools help simulate virtual users and execute test scripts. As the software market is full of various options ranging from open-source to premium, it’s difficult to find the best load testing tool for your business. In this article, we will share our list of top software load testing tools and take a closer look at their pros and cons.

    4 min read

    Website Load Testing Tutorial for QA Engineers

    website load testing with pflb preview
    Sep 8, 2025

    Performance testing is no longer reserved for specialized engineers. In the AI era, QA professionals can extend their functional testing skills to include load testing because new tools automate the most challenging aspects. PFLB offers an AI‑powered load testing platform that eliminates manual scripting and data analysis, letting you build realistic tests and receive detailed […]

  • Be the first one to know

    We’ll send you a monthly e-mail with all the useful insights that we will have found and analyzed