MapReduce Scheduler Using Classifiers for Heterogeneous Workloads

Hadoop is a large-scale distributed processing infrastructure, designed to efficiently distribute large amounts of work across a set of machines. Scheduling of jobs or work is very important in order to achieve efficiency. The proposed work incorporates the design of a new scheduler that will not overload any TaskTracker at any time and thus preventing the unnecessary re-launch of tasks. The scheduler also classifies the job into CPU bound and IO bound. So, a balance is maintained between either of them.