Created attachment 34923 [details] Aggregate report Samples Hi in the use of JMETER3.2 to do continuous stress test to see the aggregate report of the "samples" value found to be jumping growth For example: 1 5 20 40 similar to this irregular growth, but before the jmeter version is a sequential growth, for example 1 2 3 4 5, this situation in the test may cover up the server real bug
(In reply to blue from comment #0) > Created attachment 34923 [details] > Aggregate report Samples > > Hi in the use of JMETER3.2 to do continuous stress test to see the aggregate > report of the "samples" value found to be jumping growth For example: 1 5 20 > 40 similar to this irregular growth, but before the jmeter version is a > sequential growth, for example 1 2 3 4 5, this situation in the test may > cover up the server real bug Hello, I am not sure to understand your bug report. Aggregate report aggregate Sample Results, so order is not important in that regards. The change you see is related to improvement to improve responsiveness when using GUI during a Load Test which is an anti-pattern. So I'll close this bug as WORKSFORM unless I misunderstand the issue, in that case please reopen it with more details. Thanks
Hello I would like to describe the details of this problem, you can design a very simple test scene, the number of thread parameters 1, plus 1, the loop count Forever ， you will find jmeter3.2 summary report "sample" value is Jumping growth, but in jmeter3. 1 it is very smooth and is incremental, I know this may not affect the final result, but it does and the previous version is not the same, I would like to know what I mean :)
(In reply to blue from comment #2) > Hello > I would like to describe the details of this problem, you can design a very > simple test scene, the number of thread parameters 1, ramp-up 1, the loop count > Forever ， you will find jmeter3.2 summary report "sample" value is Jumping > growth, but in jmeter3. 1 it is very smooth and is incremental, I know this > may not affect the final result, but it does and the previous version is not > the same, I would like to know what I mean :)
(In reply to blue from comment #2) > Hello > I would like to describe the details of this problem, you can design a very > simple test scene, the number of thread parameters 1, plus 1, the loop count > Forever ， you will find jmeter3.2 summary report "sample" value is Jumping > growth, but in jmeter3. 1 it is very smooth and is incremental, I know this > may not affect the final result, but it does and the previous version is not > the same, I would like to know what I mean :) Hello, Sorry I still don't understand. Can you provide a test plan in English and describe what you would expect as behaviour of Aggregate Report in 3.2 and what you get ? As I wrote, the aim of our change is to make UI more responsive by updating Aggregate Report on a regular basis. This makes test much more realistic as UI update does not impact anymore load injection. Thank you
Created attachment 34929 [details] jmeter test For debugging
I uploaded the test plan, you can run in jmeter3.2 and jmeter3.1, and then you can see the difference in the aggregate report（samples）
It's important to note that in order to facilitate the test, I added the test script (jp @ gc - Dummy Sampler)
As of version 3.2 JMeter aggregates ui changes over a small period of time. The default period is 500ms. If that is too long for you, you can change it using the following property: jmeter.gui.refresh_period You can set it in bin/jmeter.properties to a lower value. Keep in mind, that lower values will result in higher CPU usage.
Thank you for your answer, solved my question! :) （This configuration is the new version of the only）
This issue has been migrated to GitHub: https://github.com/apache/jmeter/issues/4365