列存储数据库关键技术综述_李超.pdf
文本预览下载声明
37 12 V ol.37 No.12
2010 12 Computer Science Dec 2010
(清华大学信息技术研究院 北京100084)
随着互联网技术的发展、硬件的不断更新、企业及政府信息化的不断深入, 应用的复杂性 求越来越高, 推动
着数据存储技术向着海量数据、分析数据、智能数据的方向发展, 以便为数据仓库、在线分析提供高效实时的技术支
持。 基于行存储的数据库技术面临新的问题, 已经出现了技术瓶颈。 近些年来, 一种新的数据存储理念, 即基于列存
储的关系型数据库(简称列数据库, 下同)应运而生。 列数据库能够快速发展, 主 原因是其复杂查询效率高, 读磁盘
少, 存储空间少, 以及由此带来的技术、管理和应用优势。 对列数据库技术的基本现状、关键支撑技术以及应用优势进
行了介绍和分析。
列数据库, 列存储, 数据压缩, 延时物化, 成组迭代, 不可见连接, 数据仓库, 商业智能, TPCH
TP391 A
Survey and Review on Key Technologies of Column Oriented Database Systems
LI C ao ZHANG Ming-bo XING C un-xiao HU Jin-song
(Researc Institute of Info rm ation Tec nology, Tsing ua University, Beijing 100084, C ina)
Abstract Column-oriented database is a kind of new database storage tec nology t at stores data according to column
(not traditionally row).T e database pioneers suc as Dr.Mic ael Stonebraker are advocating and exploring t e new
t eory and tec nology for column-oriented database.T e main features of it are good query efficiency, less disk access,
less storage, and significant improvement of database performance.Column-oriented database is an ideal arc itecture for
data w are ouse natively, and t us s ow s a good potential in supporting ig ly efficient business intelligence applica-
tions.T is new tec nology is promising in bot academic and business, t erefore attracting lots of ig -tec corpora-
tions and researc institutes to devote in it.T is paper introduced and analysed t e main features, key tec nologies and
current R D situations of column-oriented database.
Keywords Column-oriented database, Compression, Block iteration, Late materialization, Invisible join, Data w are ouse,
Business intelligence, TPCH
显示全部