Question

Consider a web page that consists of one base html file and 20 jpg images. The base html file and 5 of the images are hosted at www.ucf.edu (UCF web server) and the other 15 images are hosted at www.google.com (Google Inc. web server). Assume that both the base html and 20 images are small enough to fit in one TCP segment each. Also, assume that our web browser is configured to have only one TCP connection at a time (i.e., no parallel TCP connections).

1. How many RTT does it take to fetch all the required files (including the base html and 20 images) if your browser uses non-persistent http.

2. How many RTT does it take to fetch all the required files (including the base html and 20 images) if your browser uses persistent http.

Answer 1

RTT: It is basically the time acquired by a small packet to travel from client to server and get back.

For Non-Persistent HTTP

[ Here for each image 2 RTT are required one for TCP connection and one for image to send.
So transmit time for 20 images = 2*(20 RTT) = 40 RTT

Total time = 2 RTT+40 RTT = 42RTT ]

For Persistent HTTP without pipelining

[ Here TCP connection is required again and again.
So for 20 images it requires -> 20 RTTs
Total time = 2 RTT(for basic html) + 20 RTT = 22RTT ]

Persistent HTTP with pipelining

[ 2 RTT for the base file + one RTT for all 20 embedded images = 3 RTT ]