I am wondering about the overhead of assertions. I created a test plan with two controllers. Each controllers runs the same set of HTTP requests, except that in one controller, each of the HTTP requests has several Response Assertions. I ran the test plan with the non-asserted controller first, then the asserted controller and the times were vastly different with the asserted controller being significantly slower.
I disabled the assertions in the one controller, and reran the test plan. Again the times were vastly different with the disabled assertion controller being slower. I swapped the controller, making the disabled-asserted controller first and reran the test. It appears that whatever controller is first, executes faster! At this point I gave up the experiment. What I would like to know for certain, is the cost of a assertion on JMeter overhead and the affect on timings. I need to use assertions to catch application errors as more concurrent users are added. But if this impacts the results significantly, I would avoid the overhead of asserts. Can perhaps one of the authors comment on the expected cost of assertions? Thanks. Scott --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
