Running JMeter Tests with Loadium
The first and second parts of the series briefly explained how to install JMeter and an example JMeter scenario. In the third part, namely this part, it will be explained how to run JMeter Tests with Loadium. Let’s start.
So, this is the first screen you’ll see once you are logged into your account. As you can see, there are 5 different testing option for you to choose. First one is JMeter Test which we will dive deep into soon.
Second option is Gatling Test. It’s also a load test framework that use Scala language. With HTTP Test option you can create a load test from scratch, you can provide URL’s, add header and body parts and define timeout values. With Webdriver Test option you can create load test scenarios based on Selenium.
Fifth option acts like a conversion tool named Convert To JMX. You can upload Postman json, HAR and then you can convert them into JMX to perform a load test.
We will be proceeding with JMeter Test option so let’s click on it. The first thing we should do is to give our test a name. If you want your tests to be more organized and accessible, you can assign them into projects as well.
From Upload File section, you can upload all your tests related files like JMX, CSV files, JAR files, etc.
Loadium supports different versions of JMeter. So, if you are more comfortable with specific version, it’s all up to you.
Configuration panel is the heart of a test creation process. Because, here we decide about the load we want to apply on our system. One of the most important value is Thread Count and it’s based on two value below. Engine Count and Thread Per Engine Count.
Engine Count means how many load generators engines will be initialized and Thread Per Engine Count means how many virtual users will be initialized for each of your engines. Engine count multiplied by Thread Per Engine Count gives you the Total Thread Count. Let’s say we have 1 engine and 5 users.
Ramp-up Time is the time for Loadium to initialize all the threads in the given time.
Iteration means the number of execution of the load script by your threads. If it’s infinite, it will keep going until the duration minute is reached.
Loadium works with Amazon Web Services and the load will be generated by Amazon Cloud Engines. For these engines, we have different geolocations for you to choose like different parts of US, Asia, Europe.
If you want your test going to be fail on specific conditions, you can define some failure criteria as well. For example your total requests’ errors percent greater than %50 stop test.
After we set all these settings, we can run the test. AWS servers initializing takes 2-3 minutes.
On the Overview page, we can see our configuration settings and also start time and run time values. In the Test Stats section, we can see values such as max users, average throughput, errors, average response time, etc.
- Max Users → Total Thread Count
- Average Throughput → Number of requests per second
- Average Response Time → Response time refers to the amount of time it takes for a server to respond to a client’s request.
The final part of this blog series will show you how to interpret the results of your tests and which parameter tells you which story. Keep reading this series for further information.