Traffic Analysis of Content Delivery Network

Content delivery network (CDN’s) have been developed to overcome the fundamental limitations of internet in terms of Quality of Service (QoS). A CDN duplicates the content of origin server to surrogate servers placed across the globe to deliver the contents to the end users in an efficient way. Content delivery on the web has received considerable research attention and the idea is to suggest an efficient networking infrastructure and replica management scenario through performance analysis of CDN. Despite its inherent limitations, even a not very complex infrastructure can show many characteristics about its ability to serve website and media traffic in a low latency model. Server utilization issues, request failures as well as mean response time are analyzed by testing different CDN policies. An efficient infrastructure is then suggested to provide guaranteed QoS for web contents using content delivery network. From this paper the readers can learn about four scenarios of CDN which are closest surrogate, Random surrogate, Load balancing surrogate and closest origin server. The scenarios are examined on the simulator od CDN where the virtual traffic passed from them to observe their response time, aborted requests and server utilization stats, with the help of generated data from the simulation results the results have been concluded.