Performance Analysis of Sales Big Data Processing using Hadoop and Hive in Cloud Environment

Abstract

Nowadays, big data gains much attention from academics and IT industries. This is due to the extraordinary current growth of data that must be accompanied by a variety of qualified data storage and processing techniques to overcome the5 V’s challenge of big data. This research is aimed to conduct a performance analysis of big data processing. The sales data will be processed in a parallel scheme on the cloud server and then managed using Hadoop and hive. The research shows that the more VMs used, the lower processing time needed, but this is inversely proportional to the CPU time required. Whereas, from the side of block size testing the research result shows that the decrease in the time of query execution is very visible by the change in the use of block size from 2MB to 4MB and 8MB, but the change in the blocksize size from 4MB to 8MB does not significantly affect the speed of query execution.