Abstract: This paper discusses Cloud Computing is rising as a replacement machine paradigm shift. Hadoop-MapReduce has become a robust Computation Model for process giant knowledge on distributed commodity hardware clusters like Clouds. Studies describe that Hadoop implementations, the default first in first out scheduler hardware is accessible wherever jobs are scheduled in FIFO order with support for different priority primarily based schedulers also other defaulter schedulers . During this paper we have studied numerous scheduling enhancements in a scheduling techniques a brand new with Hadoop like Fair4s scheduling algorithm with its extended functionalities allows processing large as well small jobs with effective fairness without starvation of small jobs. Majority of small jobs are available till date so Job scheduling algorithm must take in consideration of small jobs first with effective major to process these jobs.

Keywords: Cloud Computing, Hadoop, HDFS, MapReduce, Schedulers