We are switching to using WSGI daemon processes for different Apache virtual hosts on the same server and would like to assess the performance impact of the change.
Ideally, we would like to have web-server specific performance/load testing results and application level performance/load testing results.
Up until now, we've only used locust
and a couple specific experimental load testing scenarios.
How do you usually test this kind of web server configuration change impact? What tools do you use for web server and application performance testing? What metrics should we take into account and present as results?
One of the quick and "free" ideas we currently have is to see how an end-to-end test battery execution time will change for this environment.