Did any one use a script for a page refresh alone as a part of their stress test load profile? if yes, please let me know if I am going wrong in my case.
My case: I have a cart app on which at the end of a case, a three step buy functionality has to be stressed. I feel that due to the poor response times that users see at a high load, they tend to refresh the page at any of the 1st or 2nd stage of the buy process. But, my problem is, how much refresh is enough? I am taking it as 50% of the active load. That is the peak user load for a normal test. Should I go high? Go low? I didn't see any problem with the middle/back end with spike test which had spikes of about 150% of normal user activity.
Re: Page refresh as a part of stress test load model
Since you are testing the performance of the system I would not try to do a bunch of refreshes. Why are the TRS so high on the site. Once that issue is resolved then I don't think the users will be forced to try and refresh as much which will solve the problem. Also I am sure that the users are refreshing when the next page is only half loaded. So to do that with loadrunner I am not sure how you could do that. Would be interested in hearing about how you can do that. If you have to do these refreshes I would suggest maybe doing only 25% of the load but it really depends on how your script is layed out and the metrics you have on this refresh rate from users.
I would suggest using something like firebug when you run your test to see what is hanging on the page.