Survey on Job Schedulers in Hadoop Cluster
IOSR Journal of Computer Engineering · 2013
Hadoop-MapReduce is a powerful parallel processing technique large for big data analysis ondistributed commodity hardware clusters such as Clouds.For job scheduling Hadoop provides default FIFO scheduler where jobs are scheduled in FIFO order.But this scheduler might not be a good choice for some jobs.So in such situations one should choose an alternate scheduler.In this paper we conduct a study on various schedulers.Hopefully this study will stimulate other designers and experienced end users understand the details of particular schedulers, enabling them to make the best choices for their particular research interests.