数据流上基于遗传算法的并行查询优化技术

Parallel Query Optimization Techniques on Data Streams Based on Genetic Algorithm

传统数据库中存储的是相对静态的记录集,这些记录没有预先定义的时间概念,除非时间属性被显示地加上去.虽然这个模型能够较好地表示商业数据库和个人信息存储库,然而它对快速变化的数据流进行在线分析的支持存在很多限制.因此,需要对已有技术进行扩展研究,构建出新的管理系统来管理数据流.数据流的高速性和无限性以及计算机资源的有限性使得提高数据处理速度成为数据流管理系统(DSMS)的关键;本文主要讨论了DSMS的核心技术———查询优化;着重研究了在shared-nothing机群并行系统中,通过并行查询处理技术来提高数据流处理速度的新方法.

There is no pre - defined notion of time in traditional databases store sets of relatively static records, unless timestamp attributes are explicitly added. While this model adequately represents commercial catalogues or repositories of personal information, many current and emerging applications require support for online analysis of rapidly changing data streams. It motivates research to augment existing technologies and constructs new systems to manage streaming data. Because data streams are always high - speed and unbounded while computer resources are limited, the key problem is how to improve the data processing speed, This paper mainly discusses query optimization techniques in DSMS, with an emphasis on parallel query optimization techniques under shared- nothing cluster environment.