III-COR-Medium: Design and Implementation of the DBO Database System

III-COR-Medium：DBO数据库系统的设计与实现

• 批准号: 0803511
• 负责人: Christopher Jermaine
• 金额: --
• 依托单位: University of Florida
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2008
• 资助国家: 美国
• 起止时间: 2008-08-01 至 2010-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0803511&HistoricalAwards=false
• 关键词: III; COR; Medium; Design; Implementation

Commercial and government entities now spend around $10 billion peryear on software and hardware systems for managing "data warehouses",which are very large electronic data archives. Despite the size andimportance of this marketplace, existing data management solutions canbe painfully slow (see, for example, www.tpc.org/tpch/ for recentbenchmarking results). It is now possible to spend millions of dollarsof hardware and software for a system that still takes hours to answersimple analytic questions. This is unfortunate, because there is muchknowledge to be gained by interactive exploration of electronicarchives. Very long processing times make it likely that the data willstored away, and never looked at again.The DBO Database System project is concerned with the design anddevelopment of a unique system called DBO. Like traditionalrelational database systems, DBO can run database queries from startto finish and produce exact answers over very large archives. However,unlike any existing research or production system, DBO uses samplingalgorithms to produce a statistical estimate for the finalquery answer at all times throughout query execution. An example ofthe sort of estimate produced by DBO is, "There is a 95% chance that thetrue answer is between $1.75 million an $1.80 million". The longer auser waits, the more accurate the estimate becomes. The potentialbenefit of such an estimate is that a user can stop execution wheneversatisfied with the accuracy of the estimate, which may translate to dramatictime savings during exploratory processing. In this way, the goal ofthe DBO project is to render interactive data analysis a reality, evenover the largest databases.All scientific and technical materials produced by the project, aswell as any software available for download, can be obtained fromhttp://www.cise.ufl.edu/~cjermain/DBO .

商业和政府实体现在每年在用于管理“数据仓库”（非常大的电子数据档案）的软件和硬件系统上花费约 100 亿美元。尽管这个市场规模很大且很重要，但现有的数据管理解决方案可能速度慢得令人痛苦（例如，请参阅 www.tpc.org/tpch/ 了解最近的基准测试结果）。现在可能要花费数百万美元的硬件和软件来安装一个仍然需要几个小时才能回答简单分析问题的系统。这是不幸的，因为通过电子档案的交互式探索可以获得很多知识。很长的处理时间使得数据很可能被存储起来，并且再也不会被查看。DBO 数据库系统项目涉及一个名为 DBO 的独特系统的设计和开发。 与传统的关系数据库系统一样，DBO 可以从头到尾运行数据库查询，并针对非常大的档案生成准确的答案。然而，与任何现有的研究或生产系统不同，DBO 在整个查询执行过程中始终使用采样算法来生成最终查询答案的统计估计。 DBO 做出的估算的一个例子是，“真实答案有 95% 的可能性在 175 万美元到 180 万美元之间”。 用户等待的时间越长，估计就越准确。 这种估计的潜在好处是，只要用户对估计的准确性感到满意，就可以停止执行，这可能会在探索性处理期间节省大量时间。这样，DBO项目的目标就是使交互式数据分析成为现实，即使是在最大的数据库上。该项目产生的所有科学和技术材料，以及任何可供下载的软件，都可以从http://www.DBO.com获得。 cise.ufl.edu/~cjermain/DBO 。