论文部分内容阅读
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.
According to the multi-server architecture in distributed virtual environment, the scale expansion problem is transformed into the optimal assignment problem of multi-node and multi-server, and the evaluation equation of node classification scheme is proposed. The evaluation equation considers the system load and system delay comprehensively to achieve According to the practical application of distributed virtual environment, the node classification is divided into two steps: pre-allocation and real-time adjustment.Pre-allocation through density-first strategy and greedy adaptive search algorithm can realize fast, Effectively divide and evaluate the evaluation results by using the evaluation equation, adjust the conditions of strict restriction of node allocation in real time, and achieve the goal of real-time allocation.Compared with the traditional methods, it is a good deal of the contradiction between real-time and completeness of distribution , Has higher practical value.Through the comparison of experiments, it proves the validity of the method and draws the conclusion.