摘要
介绍了无共享并行数据库中用到的几种结点故障(处理机故障或磁盘故障)处理办法,分析了这几种办法在故障情况下保证数据库可用和负载均衡方面的差别,指出了链式分布(Chained Dcclustering)技术在上述两方面的优势。
Several methods for dealing with the problems of node failure in shared-nothing parallel database systems are presented. Also is given the analysis of difference in these methods, concerned with system availability and load balance. It is concluded that the chained declustering method is superior to the others in the two respectives.
出处
《计算机工程》
CAS
CSCD
北大核心
1995年第S1期57-61,共5页
Computer Engineering
关键词
并行处理
数据库
容错
负载均衡
parallel processing / database / fault tolarence / load balance
