Resource Aware Orthogonal Projected Regressive MapReduce Lottery Load Balancing in Cloud Computing

Abstract

Cloud Computing is an internet-based network technology that provides various services and requirements to customers through online computing resources. In the cloud, Load balancing is the most significant issue that includes both hardware and software platforms for the execution of demand of the user request. Furthermore, for handling multiple user requests, load balancing is necessary. Therefore, an efficient load-balancing technique is required to optimize and ensure user satisfaction by utilizing the virtual machine's resources efficiently. A novel Orthogonal Projected Regressive MapReduce Lottery Load Balancing (PORLOB) technique is introduced for resource-efficient task scheduling with minimal Makespan and complexity. In the PORLOB technique, many cloud user requests are transmitted to the cloud server from different locations. The load balancer uses the index table for maintaining the virtual machines. The MapReduce function includes two steps, namely, map and reduce. Based on the resource estimation, the map function performs the regression analysis and provides three resource statuses of the virtual machine: overloaded, less loaded, and balanced. In the reduction phase, the load balancer uses the lottery scheduling technique to balance the workload by migrating the task from an overloaded Virtual Machine to a less-loaded VM.