基于Petri网的集群应用系统性能评价

利用随机Petri网理论给出了一个集群应用软件负载平衡系统的抽象模型,通过细化其中的本地节点处理部分对3种集群动态负载平衡的调度策略和应用系统体系结构对负载平衡系统的影响进行了分析,得到了一些对大部分应用系统的设计起到指导作用的结论.这些结论是:(1)无论是静态负载平衡还是动态负载平衡都能提高集群系统的性能,动态负载平衡会得到更好的性能;(2)在动态负载平衡算法中除了要考虑系统中最重要的等待队列--应用队列外,还要考虑数据库队列;(3)异步体系架构将任务切分到各处理子系统中,有助于将各子系统负载数据综合到负载向量中,能够更准确地衡量系统负载、提高负载平衡系统的性能,优于同步体系结构.

PERFORMANCE EVALUATION OF APPLICATION CLUSTER SYSTEM BASED ON PETRI NET

This paper represented an abstract stochastic Petri Net model of application cluster system for load balancing.We detailed the local process part of the model in order to analyze the performance of three dynamic load balancing schemes.We also discussed the impact of application system framework on the performance of dynamic load balancing systems.It is found that:(1)Dynamic load balancing has the better system performance over that obtained with static load balancing;(2)Compared to the balancing algorithms that consider only the application queue,the load balancing algorithms which consider the database queue can get better performance;(3)Compared to the synchronization framework,the asynchronism framework can get better load balancing vector by integrating data of every sub-system.