摘要
集中式数据库的架构,使得单独的数据库管理系统(DBMS)成为了整个系统的瓶颈,已经越来越不适应海量数据对I/O性能、应用扩展、计算能力的巨大需求,正是在这种情况下大规模多数据库系统被提出,而多数据库环境多级数据仓库如何快速、合理地选择路径是研究的难点。该系统搭建在中小企业B2B电子商务系统的多数据库环境之上,以Oracle技术为支撑,通过动态路径规划以空间换时间快速、合理地选择路径,找到所需访问的目标数据库,实现多级数据仓库的ETL。实验结果表明,基于Oracle实现多数据库环境多级数据仓库ETL可以适应复杂数据仓库的并行要求,大大提高数据处理能力,有效地支持海量数据分析。
Centralized Database System makes Database Management System become the bottleneck of the entire system. And becomes increasingly unsuited to the huge demand of huge amounts of data on the I/O performance, expansion of computing power, in this case, proposes large-scale database systems, and how to effectively route storage resource is the difficulty in ring of industry. The system sets up in the multi-database environment on top of the clothing sales industry SME B2B e-commerce system, makes Oracle technical as support, through dynamic path planning space for time quickly, a reasonable choice route, locate the desired target database access, to achieve multi-level data warehouse ETL. The experimental results show that, the function based upon Oracle-based ETL system can be adapted to complex data warehouses parallel requirements, greatly improves the data processing capabilities to support massive data analysis.
出处
《现代计算机》
2014年第9期54-58,76,共6页
Modern Computer
