An Innovative Name Node Fail Safe Architecture for HDFS

Abstract

Widely used by organization, in data intensive computing is Hadoop. Client applications of Hadoop are online with unpredictable growth rate requires high reliability and scalability of system. As HDFS namespace grows, it additional memory demands are satisfied by Namenode servers vertical scalability which creates performance and availability issue. This project provides a new HDFS architecture present cloud storage scheme utilizes multiple Name Nodes, has been proposed to address the issues of scalability and availability of Name Nodes.