OK, I've been through the manual and understand how to go about setting up my tests technically. The thing that I seem to be missing is how to set up tests meaningfully.
I have an application that receives packet data on a regular basis and then allows people to log in and view reports of the data. I'd like to write a jmeter test that simultaneously simulates the data coming in and the people logging in to review it. So I've created a slew of UDP packets for coming in. It is important that packets from the same source come in sequentially. When done I will have 250 sources for these packets. I started setting up a thread group for each of the sources with one thread apiece, but it appears that jmeter wouldn't run more than 4 of them. So now what do I do to simulate this load? Should I use threads? How many, because the manual warns about too many. Then how do I set up each of my tests. I want 50 users and each user will own the data from 5 of the data sources. How do I set this part up? Thanks.