Design and Implementation of HDPS Scheduler in Hadoop Over Rackspace Cloud Server for Better Management of Data in Heterogeneous Networks

Abstract

Cloud Computing is client-server architecture based on the web services. Cloud Computing involves the process of sharing and completing the work over the internet. In this research, an optimized scheduler called HDPS (Heterogeneous Dynamic Priority Scheduler) is proposed in the Rackspace cloud server which is used to manage the task. The HDPS scheduler is used in the Rackspace cloud server to allocate and schedule a job. When a request for a job is raised from client, that task will be sent to the Cloud server for processing. Hadoop, HDFS (Hadoop Distributed File System) and Hadoop Cluster are used in this proposed work to store the files, because HDFS stores both the structured and unstructured data. Hadoop cluster normally clusters the set of data. So, an attempt has been made to implement a scheduler on the Rackspace server to work in heterogeneous environment. Hadoop and HDPS Scheduler on the Rackspace server results in reduction of the job traffics and reduce the waiting time to complete the given task.