摘要
针对监控系统数据库中历史数据表数据量大且需要频繁处理的特点,确定了对此类表进行分区优化处理的方案.阐述了表分区的优化处理方法的实现过程,将大数据量表分区存放于不同文件组,通过分散访问各个分区表,减少每次检索的数据量.最后通过大量的测试数据,对比了同一张大数据表按月份分区优化前后查询效率的变化.测试结果表明,表分区之后执行同样查询指令所需时间比分区前明显要少.查询数据时间段所跨分区越少,性能改善越明显.说明合适的表分区优化,对数据库性能有着明显的提升作用.
The realization process of data table partition optimization method is expounded. Partitions of large data table are stored in different file groups. By dispersing the visit to each partition table, the amount of data visited in each data retrieval is reduced. Finally through a large number of data test, the query efficiency of a same table before and after zoning optimization are compared. Results show that the table after zoning optimization costs less time to execute the same inquiry commands. When the time span of querying data contains less zonings, the performance improvement is more obvious. It shows that appropriate data table zoning optimization has obvious ascending effect on performance.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2011年第B09期130-133,共4页
Journal of Southeast University：Natural Science Edition
基金
江苏省徐工集团院士工作站资助项目(BM2009562)
江苏省重大科技成果转化资助项目(BA2008042)
关键词
数据库
分区优化
监控系统
monitor system
database
zoning optimization
