A use case of Content Delivery Network raw log file analysis

Abstract

The growth of video streaming has stretched the Internet to its limitation. In other words, the Internet was originally devised to connect a limited number of computers so that they can share network resources, so the Internet cannot handle a large amount of traffic at a time, which leads to network congestion. To overcome this, CDNs are built on top of the Internet as an overlay to efficiently store and swiftly disseminate contents to users by placing many servers and data centers around the globe. The topic of CDNs has been extensively studied in the last several decades. However, there is still a certain gap between theories in academia and current technologies in industry. In this paper, we take a close look at the design, implementation, solution, and performance of a CDN system by analyzing its raw log files. Specifically, its infrastructure and system design are first presented, and then we conduct a tracebased study to understand user access patterns, the sources of requests, system performance, and how such information can be used to improve the whole CDN system.