Increasing resource efficiency and reducing the energy consumption of cloud data centers is critical, especially during the global CORONA virus pandemic.Virtual machines' consolidation using live migration maximizes the hosts' and the reduction of energy consumption. An increase in the host's virtual machines in the consolidation process and the dynamic workload of the virtual machines may cause the overloading in the hosts. One approach to overcome this problem is reducing the hosts' virtual machines. One crucial issue to improve the quality of the consolidation process's quality is determining the best virtual machine for the migration process. Although the selection process has lower computational complexity than other challenges (like placement and overload prediction) in the consolidation process, this issue has received less attention. This article aims to present an efficient algorithm for the selection process. We first considered five main criteria for the selection process: migration time, migration risk, virtual machine connectivity, releasable resources, and penalty for SLA violation. Then, we propose an algorithm based on analytic hierarchy process multi-criteria decision-making technique. Next, to determine the weight of the proposed criteria, we simulate thousands of virtual machines of the PlanetLab workloads. These weights are tunable based on the data center preferences. The results of the suggested approach results show 23% reduction in the hosts' energy consumption, 49% reduction in the number of migrations, and 18% reduction in the SLA violation compared with other techniques. So, using the proposed method may significantly reduce the overall cost of the data centers.
K E Y W O R D Sanalytic hierarchy process, cloud computing, dynamic consolidation, green cloud, MADA, virtualization
INTRODUCTIONRecently, organizations, research centers, and businesses paid more attention to the advantage of using cloud data centers. These organizations prefer to use the cloud data centers to invest in private data centers and reduce the challenges and cost of their data centers. 1 So, many large companies like Amazon, Google, and Microsoft expand their large cloud data centers, which contain thousands of hosts. One of the main concerns of the data centers is the large amount of energy 1216