An Energy-Aware Virtual Machine Migration Algorithm

Virtualization is being widely applied across Information Technology (IT) infrastructure due to its various advantages including efficient resource utilization, ease of management, reduced energy consumption and much more. A Virtual Machine (VM) is a software implementation of a machine that is portable and hardware independent. VM migration improves the performance of the overall system by effectively balancing the load across servers. However, energy consumption is not considered in most of the migration algorithms. In this paper, we propose an integrated algorithm that considers the load of the server and effectively migrates the VMs. The number of VM's and various other important factors are used in selecting VMs to migrate such that the energy consumption is minimized by switching off underutilized servers. This reduces the energy cost, including the cooling cost at data centers. We evaluate our proposed algorithm using a custom built simulator. Our evaluation shows that the proposed method can reduce energy consumption by up to 20.8 % for static VM load and up to 22.0 % for dynamic VM load against pure performance based VM migration.