An Energy Efficient VM Selection Using Updated Dragonfly Algorithm in Cloud Computing

Abstract

Cloud computing is popular among industries, academia, and government to supply reliable and scalable computational power. High speed networks in cloud data centers connect Virtual machines with Physical Machines. Virtualization assists the cloud service providers to manage resources effectively but unoptimized and inefficient services degrade the performance of the system. The scheduling architecture of cloud computing includes Physical Machines (PMs), Virtual Machines (VMs) and the allocation and migration policy of the VMs over the PMs. The overutilized PMs get a few migrations and this paper introduces a novel behaviour of VM selection from overutilized PM using Swarm intelligence. The evaluation of the proposed algorithm architecture is compared with other state of art optimization algorithm from the same series. The evaluation has been done on the base of Quality of Service (QoS) parameters such as SLA-Violation, energy consumption against various load variation scenario to support elasticity. The proposed work has outcasted other techniques with significant margin in terms of QoS and the illustrations are discussed in the result.