Optimizing Make Span and Total Completion Time Together to Improve the Performance of the Map Reduce Workloads

Abstract

MapReduce and additionally Hadoop are utilized to manage bunch preparing for slots submitted from various clients (i.e., MapReduce workloads). In spite of numerous exploration endeavors committed to enhance the execution of a single MapReduce work, there is moderately little consideration paid to the framework execution of MapReduce workloads. Therefore, this paper to improve the performance of MapReduce workloads, we proposed a dynamic job ordering and map/reduce slot configuration algorithms and these to algorithms can mitigate the makespan as well as total completion time of the scheduling process of jobs. Makespan and total completion time (TCT) are two key performance metrics. Therefore, in this paper, we aim to optimize these two metrics.