2017 IEEE 37th International Conference on Distributed Computing Systems Workshops (ICDCSW) (2017)
Atlanta, Georgia, USA
June 5, 2017 to June 8, 2017
Hybrid cloud-based deployment is a trend in cloud computing which enables enterprise to benefit from cloud infrastructures while honoring privacy restrictions on some services. Enterprise application migration is an effective way to improve the efficiency of using the cloud infrastructures. However, it is a challenging problem to decide which parts of the applications to migrate and where to migrate. In this paper, we focus on the problem of planning the migration of enterprise applications in hybrid cloud infrastructures. Unlike previous studies, we consider a general hybrid cloud architecture that involves multiple public clouds rather than only one. Our aim is to maximize the enterprise cost reduction under the constraint of user experience in terms of response time. We first formulate the application migration problem as an optimization problem. Aware of its NP-hardness, we design an efficient migration framework to approximate the optimum for a large problem size. First, we leverage the application characteristic to reduce the scale of the problem by dividing it into multiple smaller subproblems. Then, an efficient algorithm based on dynamic programming is proposed to solve the small scale subproblems. Finally, we construct a feasible solution to the original problem. Simulation results demonstrate that our framework can bring significant benefits to enterprises.
Cloud computing, Servers, Optimization, Time factors, Heuristic algorithms, Computer architecture
B. Zhou, F. Zhang, J. Wu and Z. Liu, "Cost Reduction in Hybrid Clouds for Enterprise Computing," 2017 IEEE 37th International Conference on Distributed Computing Systems Workshops (ICDCSW), Atlanta, Georgia, USA, 2017, pp. 270-274.