The research for storage scheme based on Hadoop

With the rise of big data, people begin to focus on storage of it. In this paper, first, we design four different storage solutions on Hadoop platform with HBase, MySQL, XML and plain text for specific data storage. Data is stored in Hadoop Distributed File System except for MySQL storage solution, in which data is stored directly in the Linux File System. Second, for each solution, we write specific program to meet the needs. Then we run each program respectively, record execution time and monitor system performance - CPU usage, memory usage and CPU load during the whole execution. Through analyzing and comparing the differences among storage solutions, we finally conclude the advantages, disadvantages and features of each one. The results can help us to select the most appropriate storage solution with these characteristics when processing data with Hadoop platform.