Hadoop Framework to Provide Fault Tolerance in the Cluster

With the vast increase in the amount of information, nearly as high as 90% jump in volume of data compare to previously recorded values only in the past two years.As the data size gets increased, it is essential to have proper facilities to handle these data.Therefore, most of the companies use Hadoop in their application.Hadoop is an open source software that is used for reliable, scalable and distributed computing.Yet Hadoop has its own problems in its architecture which results in point of failure in the job tracker.In this paper, we design a solution to handle the problems that may occur if the job tracker fails.The architecture consists of one job tracker and several task trackers within the cluster.A monitoring agent is assigned to monitor the functionalities of job tracker and replace it with the highly efficient task tracker.The newly assigned job tracker can enhance the knowledge from the backup job tracker.