摘要： 多连接查询优化是提高数据库性能的关键问题之一。Chiang Lee提出了一种启发式多连接查询优化算法MVP，分析发现该算法并没有考虑减小执行计划的计算代价。该文结合哈希过滤的特点提出了一种改进的多连接查询优化算法，与MVP算法相比该算法降低了执行计划的计算代价，从而使查询响应时间更短。

关键词: 多连接查询, 查询优化, 图论, 哈希过滤

Abstract: Query optimization for large join queries is one of the key factors to improving the performance of database systems.Chiang Lee has put forward heuristics MVP,which finds a near optimal execution plan with complexity of O(n2).However, the computing of execution plan is not exploited. This paper presents an improved algorithm for large join queries combined with Hash filters.The algorithm can reduce more total cost of a large join query than MVP,which benefits from the following: the computing cost of execution plan is minimized.

中图分类号: 

• TP311.13