A Survey on Data Placement Strategy in Big Data Heterogeneous Environments

Abstract

In the present computerized world the extent of the information is expanding at an arbitrary speed. Simultaneously, need to process and examine huge amounts of information likewise expanded. In a few challenge commercial enterprise and logical programs, there's a want to technique petabytes of information in productive way day by day. Hadoop is most popularly used in data intensive applications. The current Hadoop usage accept that computing are homogeneous in nature. The capacity of tremendous amount of information in Hadoop is done using Hadoop Distributed File System (HDFS). HDFS employments block placement arrangement to part a really expansive record into pieces and place them over the cluster in a distributed way. In the present era of social networking, we cannot imagine having a cluster of homogeneous nodes only. Information region has not been taken under consideration for propelling theoretical outline assignments, since it is expected that the most mappings are data-local. To bargain with this Hadoop has usefulness to duplicate the information square where mappers are running. This makes a part of execution corruption particularly on heterogeneous cluster due to I/O delay or organize congestions In this survey we first brief introduction to the Big data, Hadoop and MapReduce. Later several data placement strategies in heterogeneous environment are studied.