基于随机PETRI网的云计算资源访问策略建模和分析

Cloud Computing Resources Access Strategies Modeling and Analysis Based on Stochastic Petri Nets

云计算是通过分发计算资源到大量的分布式计算机,这些计算机可以是远程的计算机或服务器,这就涉及到如何动态获取服务资源的最优方法。而Petri网和进程代数二者都是关注于系统动态行为的形式方法,因此将二者合并,可以对云计算动态获取服务资源进行性能分析,并在此分析基础上,得到资源最优选择方案。首先对云计算模型建立随机petri网模型,并对该随机petri网模型着色。其次,在文献[2]提出的SPA的基础上对该云计算的随机PETRI网模型进行性能分析,并在该基础上对该模型进行改进。最后,由于云计算的云是由多层组成的,则在云计算层次与OSPF以及petri网模型的基础上提出对服务器资源的路由选择,提出针对层次建立本地路由表、区域路由表、全局路由表,从而得到最佳路由选择方案。

Cloud computing is realized by distributing resources to a lot of distributed computers,these computers could be remote computers or servers,which relate to the optimal method that is how to obtain the dynamic service resources.Moreover,Petri nets and process algebra were both focused on the form of dynamic behavior of system,then the two methods were merged,which could be good complementary.And by the two methods,some performance analysis about obtaining service resources of cloud computing could be done dynamically,and on the basis of performance analysis,the optimal selection scheme could be got about how to get resources.Firstly,based on cloud computing model,stochastic object petri nets model was established and colored.Secondly,SPA was put forward in reference [2],and on the basis of SPA,the performance of the stochastic Petri nets model was analyzed,which was about cloud computing model,and the model based on the above analysis was improved.Finally,because the cloud computing model was made of multi-layer,routing selection for server resources was proposed on the basis of the cloud and the OSPF,and then local routing tables,regional and global routing tables were built up based on the hierarchical view.And at last,the best routing could be got.