Utilization of Data de duplication to enhance the Performance of Storage System in the Cloud

Abstract

The I/O bottleneck has become an insurmountable challenge for large-scale cloud data 
analysis due to the volatile growth in data volume. Ongoing research has shown that direct to high 
information recurrence is unquestionably present in Cloud storage frameworks. Because of the 
relatively high worldwide access area and the low number of I/O solicitations for repeating 
information, our exploratory findings reveal that information excess exhibits a much greater 
quantity of power on the I/O route than on circles. Additionally, deduplication of critical Cloud 
storage frameworks is likely to lead to memory and data inconsistencies on the platters. POD, a 
performance-oriented I/O deduplication, is proposed instead of iDedup, a limit-oriented I/O 
deduplication, to improve the I/O performance of important storage systems in the Cloud without 
sacrificing the last's limit reserves.
mentioned. Specifically, Case uses a demand-based specific deduplication procedure called Select 
Dedupe to lighten the information fracture and a versatile memory administration plot called iCache 
to facilitate the memory conflict between the bursty read movement and the bursty compose 
activity in a two-dimensional manner to enhance the execution of essential stockpiling framework 
and limit the execution overhead of deduplication. We've implemented a POD model as a module in 
the Linux working framework for your convenience. POD outperforms iDedup in the I/O 
performance test by up to 87.9 percent, with a typical of 58.8 percent, according to the tests 
conducted on our lightweight POD model. Furthermore, our findings show that POD achieves 
comparable or preferable limits on investment funds to iDedup.