Author(s): D.Sindhupriya, G.Ravi

The large utility of cloud computing has enabled enormous advances inside the real-time overall performance of systems. However, this approach results in extra storage area being used, even though reducing facts duplications would result in a decrease in information acquisitions and real time overall performance. We recommend an overall performance-oriented I/O(Input,output) deduplication(POD),instead of a capabilityoriented I/O deduplication, exemplified via iDedup, to enhance the I/O performance of primary garage systems inside the Cloud without sacrificing potential savings of the latter. POD(Performance Oriented Deduplication) takes a two-pronged approach to improving the performance of primary garage systems and minimizing overall performance overhead of deduplication, particularly, a request-primarily based selective deduplication method, called Select-Dedupe, to relieve the statistics fragmentation and an adaptive reminiscence management scheme, called iCache, to ease the reminiscence competition between the bursty study traffic and the bursty write traffic. Moreover, our assessment results also display that POD achieves similar or better potential financial savings than iDedup