Multi-object Virtual Machine Placement Optimization Method in Cloud Computing Environment

Abstract

The necessity and significance of improving the energy efficiency of cloud implementations have increased due to the rapid growth and proliferation of cloud computing services around the world. This paper has concentrated on consolidation and migration technology in virtualized cloud datacenters as a solution to deal with this high energy consumption problem. We propose VMPA strategy as a novel energy-efficient virtual machine (VM) placement procedure in cloud datacenters. VMPA could reduce the number of active servers in cloud datacenters. Besides, we introduce the central importance of optimizing different targets in cloud datacenters at the same time including total energy consumption, SLA violation and number of VM migration. We present a novel two-threshold method along with VM migration selection algorithm VMMSA as a solution of solving this multi-criteria problem. Two-threshold method can lead to less active servers and VMMSA can generate least number of VM migration. Experimental results show VMPA-VMMSA can improve energy consumption in cloud datacenters, reduce SLA violation and number of VM migration respectively in comparison with state of the arts.