Hello Adekunbi,

I have a personal preference for the Concurrency Thread Group, there are very few load testing scenarios where I had to use something else. It is really versatile and from my experience, it fits most scenarios.

About the load balancer, I think this might be a better question for the devops who set up the load balancer. What's the load balancing technique ? For example if it's a round robin, you might not have to do anything special, because the LB distributes the requests jmeter creates evenly and the responses recorded by jmeter are an average from the responses of all app servers.

Note that the requests you send will all be made from the same IP unless you use a cluster of machines to run the tests distributedly. So there might be a rule that the LB redirects traffic from the same ip to the same server. This actually happened a while ago for an app I was testing so you might have to spoof the ips (https://www.blazemeter.com/blog/how-to-send-jmeter-requests-from-different-ips).

But ideally while running the tests a person should also be monitoring the infrastructure, and if you see anomalies like all traffic being redirected to only 2 servers and 1 remains unused, you know the load balancing rules need improvement and that’s a point you can include in your report :D