Hi, My assumption is that you did something wrong in your preparation of the test and the setup if the response times are different since they're calculated the exact same way and they depend on server response time + network latency...
That aside, you should experience an actual performance hit using the open source jvm compared to the closed source kernel jvm. That implies extra CPU+RAM usage rather than a difference in MEASURED response times. The performance degradation can also include difficulties in reaching certain throughput goals (the amount of traffic generated by your test client and test plan). Still different from having actual different response times. An increase in actual measured response times can also imply A BETTER OVERALL THROUGHPUT generated by the test client using the open source JVM and this would be the opposite conclusion. You really need to make sure that you are comparing apples to apples here. I recommend Scott Barber's book on performance testing using another tool. The chapters on doing the proper measurements are tool agnostic and have helped me understand the principles that JMeter listeners also use. Cheers, -A On Sat, Aug 10, 2024 at 2:29 AM <gra...@seznam.cz> wrote: > > > " > <span lang='en' class='-wm-Y2IQFc'>is there any difference between using open > java and oracle java with jmeter? on my pc, jmeter with open java has longer > response times.</span> > " --------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@jmeter.apache.org For additional commands, e-mail: user-h...@jmeter.apache.org
