摘要
要在数据仓库环境中获得长期优良的性能最大的障碍就是发现数据仓库中大量的休眠数据 ;数据仓库中的海量数据隐藏了最终用户查询所需要的数据 ,降低了查询效率。用于提高数据仓库性能和减少休眠数据存储费用最有效的方法就是移除休眠数据。本文简要分析了休眠数据进入数据仓库的主要方式 ;改进了数据仓库中休眠数据量的统计方法 ,以便准确地计算休眠数据量的大小 ;设计了利用活动监视器监视运行于数据仓库的事务以便查找休眠数据 ;提出了用近线存储方案移除休眠数据和利用跨媒体存储器管理休眠数据的方法 。
The great obstacles are to find many idle data in data warehouse and it wants to keep its excellent performance for a long time. Many data warehouses conceal the data what the user require and reduce the query e fficiency. Removing the idle data from the data warehouse is the most effective method for improving the performance and dec reasing the fees for storaging the idle data. The main modes are analyzed for th e idle data spreading into data warehouse. In order to calculate the number of t he idle data in data warehouse the statistic method is advanced. An active monitor is used to monitor the transactions that are working in data warehouse for finding the idle data. Near line storage is used to remove the idle data and the spanning medium storage is utilized for managing the idle data . The scheme achieves a good effect in practice.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2004年第1期108-111,共4页
Journal of Nanjing University of Aeronautics & Astronautics
基金
国家"8 6 3"高技术 ( 86 3- 511- 810 - 0 4 1- 0 3)资助项目