I/O Traffic Management on PC-Cluster based Cloud Storage System

Cloud storage system architecture anddesign plays a vital role in the cloud computinginfrastructure. Cloud storage system providesusers to efficient storage space with elasticityfeature. One of the challenges of cloud storageserver is difficult to balance the providing hugeelastic capacity of storage and investment ofexpensive cost for it. In order to solve this issue,we propose the low cost PC cluster basedstorage system architecture which can beactivated to store a large amount of data andprovide cost-effective for cloud storage user andpresent FP-growth algorithm for metadataprefetching method and Least Recently Used(LRU) replacement policy for caching to improveperformance by reducing response time.According to the experimental testing, thestorage can be utilized more than 90% of storagespace and can significantly reduce the averageresponse time