Authors: Helan Liang; Yanhua Du
Addresses: School of Mechanical Engineering, University of Science and Technology, Beijing, 100083, China ' School of Mechanical Engineering, University of Science and Technology, Beijing, 100083, China
Abstract: In order to guarantee the successful execution of service processes in cloud computing, it is important to dynamically optimise service processes with temporal constraints at runtime. Recently, there exist some works related to this issue. However, they cannot adapt to mixture distributed situations where distributions of service durations are complex and diverse, and they seldom consider the adjustment penalties of service processes when compensating the temporal deficits. In this paper, an approach based on two-stage dynamic optimisation is proposed. In the first stage, we validate the temporal constraints by queuing network with considering both the uncertainty of queue time and operation time of services. In the second stage, a temporal adjustment model is designed where both temporal compensation requirements and adjustment penalties are considered, and the optimal adjustment solution is obtained by linear programming. Finally, our approach is illustrated by a real-life example in a business environment. Compared with the existing works, our approach can adapt to mixture distributed situations and obtain temporal optimisation solution with less adjustment penalties.
Keywords: service processes; temporal constraints; dynamic optimisation; queuing networks; Petri nets; cloud computing; uncertainty; linear programming; cloud services.
International Journal of High Performance Computing and Networking, 2016 Vol.9 No.1/2, pp.116 - 126
Available online: 12 Feb 2016 *Full-text access for editors Access for subscribers Purchase this article Comment on this article