| Summary: | 碩士 === 國立中正大學 === 資訊工程研究所 === 89 === In recently years, the World Wide Web has experienced phenomenal growth. This growth also contributes significant network traffic of the Internet. A popular and effective technique for improving web performance is caching. As the web cache has limited resources, it is eventually necessary to replace less useful objects with more useful ones. Several replacement policy schemes have been proposed for dealing with limited storage capacity. However, we notice that the growth rate of disk is far larger than it of memory, which will raise the impact of another reason that causes replacement - entry-constraint. We therefore propose a new replacement to fit cache system with entry constraint and a hybrid method for both disk space constraint and entry constraint.
However, a single point of web cache server has limitation on throughput and performance. Cooperative caching can therefore be a solution for this limitation. There are less previous work address issues of object placement and replacements. Hence we develop a cost model for coordinated replacement policy and replicate decisions of distributed web cache system.
|
|---|
