I was wondering if there is a a way to test the following scenario in jmeter:
Site: A site has 2 pages, one of those (let's call it heavy-page), after load is done, consumes a lot of system resources (cpu), the other doesn't (let's call it light-page).
- Simulate 1000 users (threads) loading and staying on the heavy page.
- While those 1000 users are still on the heavy-page, have another group of 1000 users load the light-page.
The key questions would be, how to have the first set of threads simulate users staying on a page after it is done loading, while having another set of users loading the light-page?
Currently I have something like this:
Group1 (1000 threads, loop count 1):
- Runtime controller (set to 5min, so each thread runs for 5 minutes):
- heavy-page
Group2: (1000 threads, loop count 1, and a startup delay set to 2 minutes, so it starts after all threads in Group1 are running)
- light-page
This does almost all I wanted, except that I see that heavy-page sent a lot more requests than what I was expecting (way over 1000) like it didn't recognize loop count set to one in G1. I'm not sure if the reason behind that is that since the sampler/thread is inside a runtime controller it is just looping forever during the time set by the controller, which would mean (I think) that the user is not just staying on the page for 5 minutes, but it is sending the request one after another after the page has loaded, which is not what I need to test
Aucun commentaire:
Enregistrer un commentaire