After following the instructions this guide to integrate tags / test run recording with a simple JMeter script, appmon tags are inserted and are processed correctly if I look at the Tagged Web Request Dashlet. However, only the last PurePath is being linked to the test results in the Test Results dashlet:
Also, regardless of how I categorize the test in the rest api call, it always shows up as a Web API test in the dashlet.
Here is the response from the API Rest Call:
Here is the x-dynaTrace header from the last purepath, correctly associated with the test run:
AN=JMeter;SI=WLT;NA=WS_LOAD_DemandeOperationCompte/v1;TR=18d2449f-6b25-4b46-a118-79869cc0a235;RC=200;VU=[VM hostname removed]_RienEntree_RienSortie_BackendNormal_0;TE=AVEC_MULTI_2018-05-25 11:53:01;SN=multiplexeurj.jmx;ID=63c821a7-6ff7-4541-a032-0201b828fe33
And here is one from another PurePath from the same run, which is correctly linked in Tagged Web Requests but not in Test Results:
AN=JMeter;SI=WLT;NA=WS_LOAD_DemandeOperationCompte/v1;TR=18d2449f-6b25-4b46-a118-79869cc0a235;RC=200;VU=[VM hostname removed]_RienEntree_RienSortie_BackendNormal_2;TE=AVEC_MULTI_2018-05-25 11:53:01;SN=multiplexeurj.jmx;ID=cc9ec29b-9303-493d-863f-54de3d433ba8
I tried with the minimum fields only as well, same behavior. We're unfortunately still at version 6.5.26 and upgrading to 7.1 this summer, but I've not seen any changes from the release notes that would affect this behavior.
Am I not sending the data correctly, or is this an appmon issue?
are you trying to do Web API testing or to run a load test? I'm asking, because I see you have selected the category "performance" for your tests (which is deprecated).
The test automation dashlet is not meant for load testing analysis. If you are trying to do load testing, since your requests correctly appear in the dashboard "tagged web requests", you can use the load test overview dashlet for analysis: https://www.dynatrace.com/support/doc/appmon/appli...
Thank you, this clarifies things, I was not aware that the performance category was deprecated, I would suggest marking it as such in the API documentation and maybe removing the corresponding section from the table in the test results dashlet.
Though if we're talking about modifications to the client, I'd rather have these tests show up in the dashlet like they do for Web API tests, as I quite like the way you can quickly chart metrics over different test runs with the drill down table / chart in the lower pane.
Does this also mean that I cannot use this for Web API testing if one tests run consists of more than one request?
Edit: also, the load test overview dashboard relies on several dashlets that require PurePath analysis, which makes it much less practical when comparing test runs that span over longer periods of time.
I thought the category had already been removed in AppMon (maybe a later version that your AppMon version, but I will check on this with R&D). Also you are right, this should be removed from the API documentation.
Yes, we recommend (and have designed!) Web API testing for only 1 request per test. Think about it as "unit testing for web service". You can theoretically misuse it, but we don't recommend it as you could end up with performance problem.
For comparing test that runs over longer periods of time, would session recording be an option for you? The workflow is described here: https://www.dynatrace.com/support/doc/appmon/continuous-delivery-test-automation/capture-performance...
Hope this helps!
Sorry for the late reply. I will have a second look at this method, but it would have been great to have the sort of quick summary / drill down on specific metrics by test runs like the Test Result Dashlet does.
Hi Carl - I have checked with the developers. I was wrong about the performance category, it is still available. I mixed it up with the "load" category, which was for load testing and was really removed from the product.
The "performance" category is basically the same as the "unit" test category (1 request per test), but you get additional metrics (time based metrics).
The rest of the info I gave you are still correct 😉