Whenever I navigate to a page in my website, 63 number of requests will hit server. Some of them will be sent concurrently and others will be sent one after the other(After receiving response from earlier request).If I record this scenario in Jmeter, club all the 63 requests in 1 Transaction Controller(because from user's point of view these all requests comprise of single action - Navigate to the page) and run the script the Aggregate report shows total response time for Transaction controller also.How Jmeter calculates this total time?
Case 1: Jmeter executes requests in series only(except for "Embedded resources" which can be run concurrently using concurrent thread pool) and it is SUM of response times of 63 individual requests? In this case the total time would be more than the actual time it takes to navigate to the page because some of the requests run in parallel in real time.
Case 2: Will Jmeter execute the few requests in parallel as a browser does? Thus total Transaction time is the duration from initial request to the final 63rd response? I believe this case gives realistic results.
But my execution reports show that Jmeter follows 1st case. Please do let me know if I am wrong.