Towards optimal resource provisioning for Hadoop-MapReduce jobs using scale-out strategy and its performance analysis in private cloud environment

Cloud computing always provides IT resources on demand basis, without additional waiting time. Therefore, data analytics is one of the most significant areas that can be benefited from Cloud Computing. MapReduce programs in the cloud computing to optimize the resource provisioning and finish the MapReduce jobs with quantified time. The efficacy as well as the accuracy of performance of the performance model based on regression used for predicting the MapReduce job completion time has been suggested in our OpenStack private cloud Hadoop cluster using linear regression method. In order to satisfy the user jobs with deadline requirements, Cloud service providers do not have a resource provisioning technique or polices. The contemporary system requires a cloud user to estimate the needed quantity of resources for running jobs in the cloud. Our proposed scalability strategy of Scale-Out methods used to obtain the accurate prediction of job completion times through our experimental results shows the performance level of MapReduce benchmark in the open stack private cloud. The regression based performance model predicting and evaluating the execution time of 5 popular MapReduce benchmark applications over our private cloud environment with better resource utilization which depicts 99% of accuracy results.