Reputation: 71
My JMeter load test includes a setUp thread group that makes 190 http requests to my test environment prior to running my test thread groups. This results in a spike of requests at the beginning of the test that appear to be included in the results of the load test. The results for average throughput are higher than they would be without the setUp thread group traffic. Is there any way to exclude the setUp thread group from the test results?
Upvotes: 0
Views: 1749
Reputation: 994
If your recording catches extra "http requests" which are not required , Click on "Add Suggested Excludes" or else can exclude by defining , so that will not come in your Result Analysis.
Upvotes: 0
Reputation: 168147
There are several options:
Synthesis Report plugin allows to set telemetry start and end offset values so you will be able to filter out unneeded data
You can add Listener of your choice to 2nd Thread Group - it will record only samplers in its scope.
You can run your JMeter tests via Taurus tool - the report will come in fancy web interface where you can zoom to the report part you need.
Upvotes: 2
Reputation: 669
Which Listener/Reporter do you use to check your results? Some reporters can show you not only average throughput but also split it by requests or threads. So, you can rename your SetUp requests using one label and just exclude these values from your total. Don't know how to do it using built-in methods but you can easily export your result table data as a file or just operate it using scripting.
If you just want to prevent spike on your graph, then use Constant Throughput Timer for your SetUp thread. It will slow down your request sending (by increasing delays between them) to meet defined throughput.
Upvotes: 0