Energy consumption of virtual machine migration within varied DCN architectures

Abstract

Virtual machine (VM) migration mechanism in data centers significantly improves the utilization of the server resource. While the recent work focuses on how to use VM migration to stability physical machine (PM) utilization or improve energy consumption, little attention has been agreed to network performance issues, such as architecture, link, load and inter-traffic between VMs in the data centers network (DCN) architectures. Network-aware Virtual Machine (NVM) placement and migration is developing as a very favorable technique for the optimization of compute network resource utilization, energy consumption, and network traffic minimization. Thus, NVM operation guarantees a fair share allocation of network resources, leading to a seamless VM mobility though decreasing degradation of network performance.The focus of this paper is to present the impact of several network architectures and virtualisation on how to reduce the energy consumed by virtual machine migration. We introduce a simulation setting for energy aware employing Green-Cloud simulator. The simulator is designed to detect the details of the consumed energy by datacenter components such us VMs, PMs, servers, switches, and links. The simulation results gotten for three-tier, Debug and High speed architectures show the value of each system in employing power management schema through DNS and DVFS by applying voltage scaling and dynamic shutdown technics.