A nalytical study on server consolidation algorithm with virtual machine migration

Abstract

Energy consumption of data centers has been increasing continuously due to rising demands of computational power. One promising approach of reducing this energy consumption is the consolidation of servers by virtualization. In a virtualized data center environment, where applications encapsulated as virtual machines are mapped to and released from the physical nodes frequently to reduce the number of nodes used. This consolidation is usually done by migrating virtual machines without stopping their services, the technology known as live migration. Moreover, server consolidation operations should be efficient, scalable and performance aware because when objects are packed together, their performance (such as throughput) degrades. So, we study on consolidation algorithms and propose an algorithm, which not only use server consolidation by minimizing the overall number of used servers to reduce energy costs, but also consider Euclidean distances to efficient placement of VMs.