期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2016
卷号：13
期号：1
出版社：IJCSI Press
摘要：Big Data is a buzz word heard everywhere and many organizations are generating huge amounts of data. The data is growing at faster pace. Variety of data stored is posing a new challenge for the organizations. Organizations need a new set of tools and techniques which can efficiently process, analyze and visualize the data for better decision making. Distributed systems developed by the developers can run on various nodes to an extent can solve the problem of data processing. Development of cloud applications is an advantage to the organizations to process the huge data on the cloud. Hadoop and its ecosystem will help to efficiently process the data by using commodity hardware. MapReduce is a framework for writing programmes for Hadoop system. Hadoop Distributed File System (HDFS) is the storage system for storing large data on commodity hardware. Hadoop file system still faces few challenges. Recently MapR has developed the MapR file system to distribute the large data sets and it overcomes the challenges faced by the Hadoop file system. In this paper we first study about the Hadoop file system, its limitations and then make a comparative study of MapR file system. Also we analyze how the MapR system is more efficient in distributing the data than Hadoop file system. We also analyze how MapR system overcomes the limitations of Hadoop File System.
