基于集群服务器的容灾系统的副本管理研究  

Research on Replica Management of Disaster Recover System Based on Cluster Server

作　　者：武鲁[1] 李伟华[1] 李钟华[1] 

机构地区：[1]西北工业大学计算机学院,陕西西安710072

出　　处：《计算机应用研究》2006年第6期76-78,共3页Application Research of Computers

基　　金：国家"863"计划资助项目(2003AA142060)

摘　　要：提出一种基于集群服务器的容灾系统副本管理方案,提出多个副本的一致性维护和副本选择的算法以及副本数量和分布方式的数学模型。通过容灾系统的性能测试实验,证明它能够实现数据的快速自动恢复,有效地管理副本,并保持副本可靠性和集群服务器性能之间的平衡。A disaster recover system replica management scheme based on cluster servers is introduced in this paper, in which a consistency maintenance algorithm of multiple replicas and a replica selection algorithm are put forward and a mathematical model of quantity and distribution of replicas is set up. Fast auto recovery of data, efficient copy management and a balance dependability of replicas and performance of cluster server can he achieved by this system scheme, which is proven by performante test of the disaster recover system.

关 键 词：容灾系统 数据恢复 副本一致性 副本选择 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]