Optimal Allocation of Cache Servers and Content Files in Content Distribution Networks

A. Nakaniwa and H. Ebara (Japan)


Content Services, Mirroring, Cache Server, CDNs, Internet


Content Distribution Networks (CDNs) have proliferated rapidly with the growth in network bandwidth and the va riety of content files. The CDN service is one of the most popular Web services from Internet Service Providers (ISPs), since most ISPs are currently striving to provide ad ditional services besides their original Internet access ser vice. The Internet has seen an explosive increase in the number of users, which causes network servers to become overloaded, and we are currently facing several serious is sues, such as declining reliability, increased response time, and so on. The replicated allocation of content files on cache servers for load-balancing is known as an effective solution to overcome these issues. In CDNs, one of the most important problems is how we allocate cache servers and content files while considering trade-offs between the cost, reliability, and delay. In this paper, we examine an example of the CDN service in an ISP ”BBit-Japan”, which is one of the biggest CDN service providers in Japan. We apply this example to the cache server and content allocation model which we have proposed. This optimization model introduces 0-1 in teger programming to determine the optimal allocation of cache servers and content files. In this optimization model, we aim to maximize the reliability of the whole system subject to restrictions of cost and delay, and in considering trade-offs between reliability, cost, and delay. The use of this optimization model makes it possible to show precisely how they can properly allocate cache servers and content files at limited cost and within permissible delays to the CDN service providers.

Important Links:

Go Back