Using Taurus for Load Tests with 100+ Labels
If your JMeter performance test contains more than 100 labels, the BlazeMeter reports will display the metrics pertaining to the first 100 labels only. Any further labels beyond the first 100 will be combined under a label named “AGGREGATED LABELS”. Likewise, a multi-test will keep track of up to 300 labels before aggregating them. This does not enable you to fully monitor the test.
Aggregated labels (See “AGGREGATED” below “ALL”)
However, several large-scale enterprises have legacy test scripts with more than 100 labels. Due to the limitation mentioned above, they have to spend several labor hours cleaning up their scripts, either in the form of reducing the labels or splitting them into smaller ones.
Taurus provides a solution. The external-results-loader executor enables you to load test results from output files generated outside of Taurus. This includes JMeter’s JTL files, Gatling’s simulation.log, etc. You can then quickly generate an online BlazeMeter report with the executor (and the output files). Therefore, by loading your JMeter jtl files to Taurus and running them in BlazeMeter, you can bypass the 100 label limitation and view the metrics at an individual label level.
Let’s see how it works.
1. Once your BlazeMeter test is complete, download the artifacts.zip folder for all the Load engines. (Or, take your .jtl files from a standalone JMeter test).
2. Collect kpi.jtl and error.jtl files from all the artifacts folders and copy/save them into a separate folder.
3. Configure a yaml file by including the data-files along with the executor external-results-loader, as shown in the example below:
- The execution section of the YAML specifically shows the input .jtl files and the use of the external-results-loader module.
- The modules section is where you configure your BlazeMeter API Key & Secret so that your report gets saved under your BlazeMeter account.
- The reporting section is where you may configure a name for your report and specify the project under which you need it to be saved within BlazeMeter.
4. Generate the report for all the labels by using the Taurus external result loader. Run the command: bzt <filename>.yml -report
5. Providing a valid <api key> will save the generated report with all the labels under the BlazeMeter account associated with the <api key>. Otherwise, the report will be generated anonymously.
That’s it! As you can see below, there are no more aggregated labels in the report:
Please note that Taurus can take the same amount of time as your test to load the metrics into BlazeMeter. You will also consume appropriate test credits (VUHs) if you attempt to save this report under your BlazeMeter account. No test credits are consumed if you want to view an anonymous report.