Abstract: Static availability of Web cluster system is gained by building the Markov model of the system state and utilizing the properties of state diverting probability balance in steady-state Markov model. The research focuses on system’s high availability of dynamic configuration. A new dynamic availability scheduling algorithm based on table are presented to realize dynamic high availability of Web cluster. The result of the experiments shows the validity of the algorithm.

摘要： 通过建立Web集群系统状态的马尔可夫模型,并利用稳态马尔可夫模型中状态转移概率平衡的特点来求解系统静态配置可用度,着重对系统动态配置的可用度进行了研究,提出了一种基于表的动态可用度调度算法来实现系统的动态高可用度,实验结果表明了算法的有效性。

关键词: Web集群, 静态可用度, 动态可用度

CLC Number: 

• TP393.02