HDFS (Hadoop Distributed File System) is a distributed user level file system which stores, processes, retrieves and manages data in a Hadoop cluster. HDFS infrastructure that Hadoop provides, include a dedicated master node called name node which contains a job tracker, stores meta-data, controls the overall distributed process execution by checking out whether all name nodes are functioning properly through periodic heart beats. It also contains many other nodes called data node which contains a task tracker, stores applications data. The Ethernet network connects all nodes. HDFS is implemented in Java and it is platform independent.