A cost-effective cloud storage caching strategy utilizing local desktop-based storage

Abstract

This paper proposed a cost-effective cloud storage caching strategy by utilizing local desktop-based storage, called CloudCache. Nearly free desktop machines in local area network environment are utilized to build a local distributed file system, which is deployed as a data cache of remote cloud storage service. When the user reads a file, it first determines whether the file is in the cache. If it is in the cache, the file is read directly from the cache; otherwise, the file is read from remote cloud storage service. Least Recently Used (LRU) algorithm is used for cache replacement. Performance evaluation is accomplished, using Amazon Simple Storage Service (S3) and desktop PCs in laboratory to build the experimental environment. A case study of web cache has also been demonstrated. The results show that using the proposed caching strategy not only reduces cost, but also greatly improves file reading speed.