大型管理信息系统(MIS)中数据库瓶颈的分析与优化

Analysis and Optimization of the Database Bottleneck in a Large-scale Management Information System(MLS)

数据库是信息管理系统瓶颈常发环节。性能调优及代码优化（软优化）虽能解决绝大部分瓶颈问题,但对海量数据的访问性能提升依然存在先天不足,而物理层面对I/O的优化（硬优化）往往能够起到突出效果。实验结果表明,针对基本表设计、表分割以及表对象放置的优化策略对海量数据访问性能的提升起到良好作用,可作为软优化的重要补充。

Database is the part where there are frequently bottlenecks in an information management system.Soft optimization methods such as capability and code optimization are able to solve most of bottlenecks,but still unable to help improve the performance of mass data visits.However,the physical level usually exerts extraordinary effects on I/O optimization （hard optimization）.Experiment results show that the optimization strategies used in basic chart design,chart division and placement of chart subjects are favorable to improve the mass data visit performance and can be used as significant supplement to soft optimization.