基于数据动态冗余的分布式并行系统重构机制  

Mechanism of Distributed and Parallel System Reconfiguration Based on Dynamic Redundancy of Data

作　　者：陈建英[1] 刘心松[1] 左朝树[1] 陈小辉[1] 

机构地区：[1]电子科技大学,四川成都610054

出　　处：《计算机应用研究》2004年第11期229-231,256,共4页Application Research of Computers

摘　　要：设计高可用分布式并行数据库系统,系统重构不能只局限于单节点状态变化,为此提出了基于数据动态冗余的分布式并行系统重构机制。该机制通过相关事件触发系统检测,及时分布采集节点状态信息,借助数据的再生和转移,在各节点并行完成负载失衡和节点状态变化下的系统自动重构。实际应用表明,该机制解决了多节点状态同时变化的系统资源和数据资源的重构问题,保证了数据库系统的高可用性。To design a high availability distributed and parallel database system,the system reconfiguration shouldn't be limi-(ted) to the change of single node state.Therefor,a mechanism of DPSR(Distributed and Parallel System Reconfiguration) based on dynamic redundancy of data is put for ward.In this mechanism,the system detection is triggered by correlative events,the information of node states is collected timely and distributedly,and the system reconfiguration is executed automatically and parallelly at each node by the way of regeneration and transfer of data.The applications of DPSR used in practice illustrate that the mechanism has sloved the reconfiguration problem of system resources and data resources under the simultaneous change of multiple node states and has ensured the high availability of the database system.

关 键 词：分布式并行数据库系统 系统重构 数据动态冗余 高可用性 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]