Why would the same load test have different results when done with LR9.1 and LR9.52?

  • Questions
  • Why would the same load test have different results when done with LR9.1 and LR9.52?
Question ID: 104161
3
0

Can Anyone Explain why a load test ran with LR 9.1 would clearly show a climbing trend in the transaction response times and then when running the same test ( nothing changed in the script or the application) ran with LR 9.52 would in not show the climbing trend?. Which test result are correct?

Please list the possible reason for this, Thank you

Marked as spam
Posted by (Questions: 12, Answers: 363)
Asked on July 22, 2010 4:55 pm
16 views
Answers (2)
4
Private answer

If you are using a protocol for which there was memory leak in 9.1 and that memory leak was fixed in 9.52, or if you have very large correlation parameters that can cause memory to get fragmented, then it is possible that LR 9.52 improved memory handling.

Marked as spam
Posted by (Questions: 1, Answers: 5)
Answered on October 19, 2010 7:52 pm
2
Private answer

No, I can't answer this. This should not happen. My only guess is that perhaps the client settings (Web HTTP?) were different between test runs, or that the browser caching settings were different.

Marked as spam
Posted by (Questions: 0, Answers: 2)
Answered on October 6, 2010 2:43 pm