High Performance Hadoop Distributed File System

Abstract

A Distributed File System (DFS) is client/server software paradigm. It gives the client’s ability to access the server data. In general, DFS includes software exist on the servers and clients that connects files on different file servers into a single Namespace to improve the data availability. In case of a client retrieving a file from the server, the retrieved file will appear on the client machine as a normal file, so the client able to access the file as if it were stored locally on the client machine. After that, client stores the modified file to the server.