Question

suppose that the web page associated with the link contains 6 small objects that are referenced by the base html file (7 objects in all), all located on the same server. let rtt 1 denote the rtt between the local host and the server containing the objects. assuming zero transmission time of the objects, how much time elapses from when the client clicks on the link until the client receives all the objects, when non-persistent http with pipelining is used? how does this change when the browser is configured for 6 parallel connections?

Answer 1

Non-persistent HTTP with pipelining takes 2 RTTs to receive all objects as requests can be sent consecutively in one connection, while 6 parallel connections also take 2 RTTs but send requests simultaneously across multiple connections.

Step-by-step explanation:

The question asks about the time elapsed when a web page with 7 objects (including the base HTML file) is downloaded using non-persistent HTTP with pipelining and compares it to a situation where the browser uses 6 parallel connections. With non-persistent HTTP without pipelining, each object requires a separate connection. This would mean the client experiences 7 RTTs (Round Trip Times) - one RTT to open the connection and request the base HTML file, and 6 additional RTTs for each of the small objects. However, if pipelining were available, all objects could be requested in the first connection, right after the base HTML; thus, one RTT is needed to setup the connection, another RTT for requesting and receiving the base HTML, and one final RTT for requesting and receiving all referenced objects, totaling 2 RTTs.

For the scenario with 6 parallel connections, each object could be requested simultaneously. The total time would be just two RTTs - one for setting up all 6 connections and another for requesting and receiving the objects.