Department of Computer Science and Engineering, Marathwada Institute of Technoloy, Aurangabad, Maharashtra, India
Online published on 8 November, 2017.
Hadoop is a new technology in parallel computing and its performance in all aspect is increasing day by day. There are various benchmark suite for performance analysis such as TeraSort, TestDFSIO, NNBench and MRBench. But in this paper we have used TeraSort as a benchmark suite for analyzing performance of different versions of hadoop and we campared the results among different versions of hadoop. Main theme behind this project is to analyze Map, Shuffle and Reduce Time on different versions of Hadoop with different input size of data ranging from 100MB to 400 MB. In these experiments we have used two nodes. Each node has dualcore intel processors and 2 GB RAM.
Hadoop, TeraSort, Map, Shuffle, Reduce, MapReduce