Abstract

Mapreduce in hadoop is parallel processing. In modern world data centres operate various MapReduce function in parallel, hence it is essential to furnish an effective scheduling algorithm in order to optimize completion time required for these jobs. Hadoop adapted FIFO scheduling as default scheduling algorithm but maybe this scheduling is not powerful to fulfil the requirements of all jobs. So in this situation one should use alternate scheduling algorithms. This study will incite other experience clients and designers to conceive the details of certain scheduling, enable them to make the best decisions for their certain research interests.