Phase Aware Job Scheduling for MapReduce in Hadoop
Snehal Kapde, J V Shinde, Nilesh R. Wankhade · IJARCCE · 2016
MapReduce is a framework using which we can write applications to process huge amounts of data, inparallel, on large clusters of commodity hardware in a reliable manner.MapReduce is a model for the dataintensivecomputation.However, despite recent efforts towards designing efficient scheduler to perform MapReduce, the available solutions focuses on scheduling at the task-level offers suboptimal performance in executing jobs.This isbecause tasks can have highly varying resource requirements during their lifetime, which makes it difficult forschedulers to effectively utilize the available resources to reduce job execution time.PRISM-a fine grained phaseand resource aware scheduler was introduce which mainly focuses on designing task level scheduler.The phasebased resource aware scheduler offers high resource utilization and provides improvement in job running time.In my proposed system I have used the virtual memory to overcome the disadvantage of PRISM and using thepausing phase.In this the resource will able to use virtual resource.Instead of going into paused phase it is possibleto use resource virtually till the resource is available.This will surely help the job running significantly reducing the delay.