基于多核环境的并行性双向枚举连接

Parallel Two-Way Enumeration Join Based on Multi-cores Environment

基于多核处理器,结合自底向上和自顶向下两种算法,提出一种图遍历驱动的双向优化算法,该算法充分利用两种遍历算法的优点,并发挥多核环境的优势,实现了最优查询计划的高性能并行构建,解决了并行双向枚举连接问题.实验结果表明,该算法的性能优于已有算法,可明显提高数据库查询速度.

Multi-j oin query optimization is one of the key problems in database searching.As multi-cores replace traditional single processor because of the advantages of their high performance and low energy consumption, researches on bottom-up dynamic query parallel algorithms have become popular.However,the advantages of top-down dynamic query algorithm can not be replaced by bottom-up dynamic query algorithms,such as cut algorithm.Based on bottom-up and top-down algorithms,this paper proposes a graph traversal two-way optimization algorithm on the multi-cores system.The proposed approach makes full use of the advantages of both top-down and bottom-up algorithms, gives full play to multi-cores environments, realizes high performance parallel construction of optimal query plans, thus resolves parallel two-way enumeration j oin problem. Experimental result shows that the proposed algorithm outperforms existing algorithms,and can effectively increase the speed of database query.