CS结构下的DVE规模扩展方法

针对分布式虚拟环境下的多服务器结构,将规模扩展问题转化为多节点与多服务器的最优分配问题,并提出了节点划分方案的评估方程.评估方程综合考虑系统负载、系统延迟,以达到对分配方案的有效评估.根据分布式虚拟环境实际应用的特点,将节点划分分为预分配和实时调节两个步骤.预分配通过密度优先策略和贪婪自适应搜索算法实现对预分配节点快速、有效地划分,利用评估方程对评估结果进行修订;实时调节严格限制节点分配的条件,达到实时分配的目的.相比于传统方法,很好的处理了分配的实时性和完备性之间的矛盾,具有较高的实用价值.通过实验比较,证明该方法的有效性,并得出结论.

Method of scale extension in distributed virtual environment under CS structure

According to the client server(CS) architecture in distributed virtual environment(DVE), it extended the problem of scale extension into the optimization of node partition between multi-node and multi-server. An evaluation equation of the node partition was also proposed. Based on the characteristics of DVE application, the whole process was divided into two steps,predisposition and runtime regulation. The process of predisposition was achieved partition rapidly and efficiently with density primary strategies and greedy randomized adaptive search procedures(GRASP), and the evaluation equation was used to adjust partition result. In real-time regulation process, the condition of node partition was restricted strictly to achieve real time distribution. Compared with the traditional methods, this method has better solved the contradiction between the real-time quality and the maturity of distribution, and has been proved to be of great practical value. The experimental results illustrate the effectiveness of this method and a conclusion was drawn.