Capturing response times for individual transaction run
I have a script that I have set its TMain transaction to run more than once. When I run 1 Vuser for that script I am getting reponse time results that seem bogus to me:
31.453000,16.687000, (TMain 1)
34.859000,28.328000, (TMain 2)
38.327000,30.140000, (TMain 3)
44.452000,35.000000, (TMain 4)
47.686000,37.907000, (TMain 5)
58.171000,47.673000, (TMain 6)
etc...(depending on what the value TMain is set to)
These results seem to increase with each repeat of TMain - why is that?
I am running a Queuing workload set to 1 Vuser and all other values to 0.
I've tried steady state and increasing workload but with similar results.
The reason I question these results is this:
I ran a 5 Vuser queuing workload and wrote the results to the results repository - none of the max results were as big as the above reponse times! - what gives?
I need to know numbers by run i.e. for any given user how long it takes to do operation A and operation B in TMain for the first, second, third, etc. time, and I would like to know if:
1. Am I approaching this the correct way?
2. If so, then why are the numbers increasing with each run?
thanks in advance.
[This message has been edited by QA_1 (edited 04-07-2003).]