Dynamic and Adaptive Replication Mechanism for Improving Bandwidth in Hadoop Distributed File systems

Abstract

HDFS (Hadoop Distributed File System) is a disseminated storage scheme that efficiently stores and streams comprehensive data sets to high-bandwidth applications. The popularity of HDFS data fluctuates over time. HDFS offers high performance, trustworthiness, & accessibility by replicating data, normally three duplicates of each data. Replication means keeping more than one system with the same copies of data. Replication is one of the most important file system design requirements (DFS).