An Evolutionary Computing based Energy Efficient VM Consolidation Scheme for Optimal Resource Utilization and QoS Assurance

Background: The increase in Cloud applications have demanded efficient cloud computing systems like Virtual Machine (VM) consolidation that intends to facilitate optimal resource utilization, energy conservation and quality of service. Methods: In this paper, an evolutionary computing technique called Adaptive Genetic Algorithm (A-GA) has been proposed for VM consolidation that encompasses under load and overload utilization detection, VM selection and placement, where the modified robust local regression and interquartile range schemes estimate the dynamic CPU utilization threshold for overload detection, minimum migration time works as VM selection policy, while A-GA optimizes VM placement across network to reduce energy consumption and SLA violation. Findings: PlanetLab Cloud benchmark data based simulation results confirms that the proposed VM consolidation scheme exhibits better than other existing approaches such as Ant Colony Optimization (ACO), Static Threshold (THR), Local Regression (LR), Conventional Inter Quartile Range (IQR) and Median Absolute Deviation (MAD) based virtualization schemes. The proposed system has exhibited minimal host shutdown, VM migration, energy consumption and SLA violation as compared to other existing approaches. Applications: Thus, the efficiency of the proposed VM consolidation scheme signifies that it can be a potential VM consolidation solution for large scale Cloud data centers.