ATan (Bug 64103):
Whenever I have a throughput controller paired with a http sampler, I get the occasional stack overflow error when trying to generate a dashboard report. I've included a basic script of what my test plan looks like, but it basically goes like this:
Thread group - 250 users
-Read value line by line from a csv file
-Include value in http payload in the form of a POST request to a server
-Throughput total executions for 1 Million requests. I want the test to gracefully end and generate dashboard after it fulfills this
More often than not I have an error at the last step. And in addition to that, during the very last step of running this test the latency has a huge spike. From an average max of ~100 ms to sometimes ~900000 ms. I even tested with a different server and its the same behavior.
And for clarity sake, I'm encountering problematic behaviors towards the end of the test when I use a throughput controller. Either the test wraps up cleanly, stack overflows, or has a huge lag spike before the report is created.
I'll try attaching the error message from log in addition to test plan. I've seen similar bug requests involving if statements, but I can confirm there are no explicit if statements in my test plan.
+Java 8
+20GB Heap memory, modified registry timedwaitdelay and maxuserport for max values
ATan (Bug 64103): Whenever I have a throughput controller paired with a http sampler, I get the occasional stack overflow error when trying to generate a dashboard report. I've included a basic script of what my test plan looks like, but it basically goes like this:
Thread group - 250 users -Read value line by line from a csv file -Include value in http payload in the form of a POST request to a server -Throughput total executions for 1 Million requests. I want the test to gracefully end and generate dashboard after it fulfills this
More often than not I have an error at the last step. And in addition to that, during the very last step of running this test the latency has a huge spike. From an average max of ~100 ms to sometimes ~900000 ms. I even tested with a different server and its the same behavior.
And for clarity sake, I'm encountering problematic behaviors towards the end of the test when I use a throughput controller. Either the test wraps up cleanly, stack overflows, or has a huge lag spike before the report is created.
I'll try attaching the error message from log in addition to test plan. I've seen similar bug requests involving if statements, but I can confirm there are no explicit if statements in my test plan.
+Java 8 +20GB Heap memory, modified registry timedwaitdelay and maxuserport for max values
Created attachment Bug.jmx: Test Plan
Bug.jmx
````xmlSeverity: major OS: