cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

This product reached the end of support date on March 31, 2021.

Test Automation Integration tests on different environments

Bert_VanderHeyd
Advisor

At our customer we are performing Integration Tests on an application with TFS using the TFS Plugin on different environments. The development environment and a more stable test environment.

Both these environments are instrumented and the test automation features of AppMon are applied. Currently they all end-up in the same system profile since they share certain services. So, in the Test Automation dashlet you see both the dev as test testruns. The only thing they have different is the version. We added the environment within the versionBuild. So I guess that they currently both contribute to the same test metrics and both influence the corridor since only the version is different.

My Question: Is there a property we can use (maybe "Platform") to get a clear separation between the tests without the need for a new System Profile?

I am pointing to the rest interface here to register the test-run:

https://dynatrace-server:8021/api-docs/current/ind...

5 REPLIES 5

Wiktor_Bachnik
Dynatrace Mentor
Dynatrace Mentor

Hello Bert,

You are right, adding environment name to the version build field doesn't separate metrics, so in the test results view you'll see both in the same chart. If you're unable to use separate system profiles, I think using the platform field would be the second best approach. That will cause all the tests to be listed separately for each platform in the Test Results view.

Please note that in the Tests Overview dashlet you can use the platform field as a filter, so you can see the overview for each platform separately (you may want to save a dashboard for that).

Best regards,

Wiktor

Perfect, I already tried it and a different platform indeed results in separate graphs for each platform.

rick_boyd
Mentor

Hi Bert,

This might be a secondary concern but to the question of whether it will affect the test corridor, each metric is automatically baselined on the basis of the agent instance (agentname@host) so in this regard it's already being split out like that

This doesn't seem to be the case. Also because the first agent, which starts the actual test, is the TFS release node in both cases.

Ah yes definitely then since the baseline is on the entry point agent