ABSTRACT

MapReduce is used for processing and generating sets large data .A open source framework of MapReduce is Hadoop [1]. MapReduce and Hadoop represent a good alternative for efficient large scale data processing and advanced analytics in an enterprise. In Heterogeneous computing, map or schedule a processor to a single core or different type of processors to a single core or a processor to many cores or many processors to many cores. So the usage of heterogeneous multi-core processors for the efficient performance in map reduce environments is increasing. Therefore the single heterogeneous multi-core processors consists of small and big cores where small cores provide power efficient and big-cores provide high-performance , which includes Inductive Logic Programming and Multilayer Perception to be extracted dynamically. This paper addresses various scheduling approaches that helps in improving the performance in heterogeneous environment. The outcome of this paper shows that traditional approaches used in Hadoop suffers from various issues. This paper will encourage in addressing those issues and discusses various scheduling approaches which helps in big data analysis.