A Cooperative Multi Agent Learning Approach to Manage Physical Host Nodes for Dynamic Consolidation of Virtual Machines

One of the most important challenges in a virtualized cloud data center is to optimize the energy-performance tradeoff, i.e., finding the right balance between saving energy and attaining best possible performance.Distributed dynamic virtual machine (VM) consolidation (DDVMC) is a virtual machine management strategy that uses a distributed rather than a centralized algorithm for finding such optimums, here also aiming at increasing scalability by avoiding a central bottleneck.The general goal of DDVMC in data centers is to (1) manage physical host nodes in order to avoid overloading and underloading, and (2) to optimize the placement of VMs.However, the optimality of this strategy is highly dependent on the quality of the decision-making process. In this paper we concentrate on managing physical host nodes in DDVMC strategy and propose a cooperative multi-agent learning paradigm to make optimal decisions towards energy and performance efficiency in cloud data centers. Our approach is also able to assure scalability due to increasing the number of hosts in the data center. The experimental results show that our approach yields far better results w.r.t. the energy-performance tradeoff in cloud data centers in comparison to state-of-the-art algorithms.