Energy-aware and multi-resource overload probability constraint-based virtual machine dynamic consolidation method

作者：Li, Zhihua^*; Yan, Chengyu; Yu, Lei; Yu, Xinrong

来源：Future Generation Computer Systems-The International Journal of eScience, 2018, 80: 139-156.

DOI：10.1016/j.future.2017.09.075

• Virtual machine consolidation
• Multi-resource constraints
• Optimization model
• Artificial bee colony-like search
• Energy aware

In a data center, virtual machine consolidation has been proposed to improve the resource utilization and energy efficiency. An effective and efficient virtual machine consolidation method should achieve an appropriate balance among multiple goals, including guaranteeing service quality, reducing energy consumption and maximizing resource utilization. This problem is a multi-objective optimization problem with multiple resource constraints. To solve this problem, we propose an energy-aware dynamic virtual machine consolidation (EC-VMC) method that migrates virtual machines while satisfying constraints on the probabilities of multiple types of resources being overloaded. In our method, a series of algorithms for selecting and placing virtual machines to be migrated are utilized, with constraints on the probabilities of various resources in a physical machine being overloaded. Our algorithms integrate and cooperate similarly to artificial bee colony foraging behavior to perform an optimized search for the mapping relation between virtual machines and physical machines for consolidation. Extensive simulation is conducted to compare our EC-VMC method with previous virtual machine consolidation methods. The simulation results demonstrate that the EC-VMC method effectively overcomes the deficiencies of some existing heuristic algorithms and is highly effective in reducing VM migrations and energy consumption of data centers and in improving QoS.

• 出版日期2018-3
• 单位江南大学