Optimization-Based Virtual Machine Manager for Private Cloud Computing

Abstract

In this paper, an optimal resource management framework for cloud computing environment is presented. Based on virtualization technology, the workload to be processed on a virtual machine can be moved (i.e., outsourced) from private cloud (i.e., in-house computer system) to the service provider in public cloud. The framework introduces the virtual machine manager (VMM) in private cloud operating to minimize the cost due to the outsourcing and performance degradation. A stochastic optimization model is developed to obtain an optimal workload outsourcing policy with an objective to minimize a cost. The numerical studies reveal the effectiveness of the optimal resource management framework to achieve an objective of private cloud. This framework will be useful not only to optimize the performance of resource usage, but also to achieve the best benefit from economic perspective of the cloud computing regime.