Virtual Machine Task Allocation for HLA Simulation System on Cloud Simulation Platform
A new yet promising technology, Cloud computing, can benefit large-scale simulations by providing on-demand, everywhere simulation services to users. In order to enable multi-task and multi-user simulation tasks with Cloud computing, Cloud Simulation Platform (CSP) is proposed and developed. To promote the running efficiency of HLA systems on CSP, this paper proposes an approach addressing the Virtual Machine task allocation problem, which is divided into two levels of task allocation steps. The first-level uses a heuristic algorithm to optimize the mapping from federates (of HLA system) to virtual machines (of CSP) and aims to achieve load balance on virtual machines in CSP. The second-level dispatches the subtasks of federate to the cores of virtual machines to minimize the makespan (schedule length) of the federate which uses a DAG based list scheduling algorithm: the EST (Earliest-Start-Time) algorithm. Experiments show that the two-level task allocation strategy effectively improves the running efficiency of HLA system on CSP.
Keywordstask allocation Cloud Simulation Platform HLA DAG heuristic algorithm
Unable to display preview. Download preview PDF.
- 1.Li, B.H., Chai, X.D., Hou, B.C., et al.: Networked Modeling & Simulation Platform Based on Concept of Cloud Computing-Cloud Simulation Platform. Journal of System Simulation 21(17), 5292–5299 (2009)Google Scholar
- 2.Topcuoglu, H., Hariri, S., Wu, M.Y.: Performance-Effectvie and Low-Complexity Task Scheduling for Herogeneous Computing. In: IEEE Transachtions on Parallel and Distributed Systems, pp. 260–274. IEEE Press, New York (2002)Google Scholar
- 3.Gary, M.R., Johnson, D.S.: Computers and Intractability: A Guide to the Theory of NP-Completeness. W.H. Freeman and Company (1979)Google Scholar
- 4.Kwok, Y., Ahmad, I.: Dynamic Critical-Path Scheduling: An Effective Technique for Allocating Task Graphs to Multi-processors. IEEE Transaction on Parallel and Distributed System, 506–521 (1996)Google Scholar
- 5.Boukerche, A., De Grande, R.E.: Dynamic Load Balancing Using Grid Services for HLA-Based Simulations on Large-Scale Distributed Systems. In: Proceedings of 13th IEEE/ACM International Symposium on Distributed Simulation and Real Time Applications (DS-RT 2009), pp. 175–183 (2009)Google Scholar
- 6.Audsley, N., Burns, A., Richardson, M., et al.: Applying new scheduling theory to static priority pre-emptive scheduling. Software Engineering Journal, 284–292 (1993)Google Scholar
- 7.Stavrinides, G.L., Karatza, H.D.: Scheduling multiple task graphs in heterogeneous distributed real-time systems by exploiting schedule holes with bin packing techniques. Simulation Modelling Practice and Theory, 540–552 (2011)Google Scholar
- 8.Du, Z.H., Wang, M., Chen, Y.N.: The Triangular Pyramid Scheduling Model and algorithm for PDES in Grid. Simulation Modelling Practice and Theory, 1678–1689 (2009)Google Scholar
- 9.Sih, G.C., Lee, E.A.: A Compile-Time Scheduling Heuristic for Interconnection-Constrained Heterogeneous Processor Architectures. IEEE Transaction on Parallel and Distributed Systems, 175–187 (1993)Google Scholar