A Review of Scheduling Algorithms in Hadoop

In this epoch of data surge, big data is one of the significant areas of research being widely pondered over by computer science research community, and Hadoop is the broadly used tool to store and process it. Hadoop is fabricated to work effectively for the clusters having homogeneous environment but when the cluster environment is heterogeneous then its performance decreases which result in various challenges surfacing in the areas like query execution time, data movement cost, selection of best Cluster and Racks for data placement, preserving privacy, load distribution: imbalance in input splits, computations, partition sizes and heterogeneous hardware, and scheduling. The epicenter of Hadoop is scheduling and all incoming jobs are multiplexed on existing resources by the schedulers. Enhancing the performance of schedulers in Hadoop is very vigorous. Keeping this idea in mind as inspiration, this paper introduces the concept of big data, market share of popular vendors for big data, various tools in Hadoop ecosystem and emphasizing to study various scheduling algorithms for MapReduce model in Hadoop and make a comparison based on varied parameters.