数据虚拟化平台的设计与实现

Design and implementation of data virtualization platform

为了解决大数据所面临的数据源多样化、数据规模大、数据异构异质性、数据抽取、共享和整合等问题,提出了一种利用数据虚拟化技术来设计和构建数据虚拟化平台以实现数据融合的方法。首先,从整体上介绍了数据虚拟化平台的设计架构;其次,详细介绍了源数据层对异构、异质数据的处理方案;再次,分析了虚拟数据对象层的结构并对数据的抽取和回写等方法进行了详细的论述;最后,阐述了应用接口层的实现机理和实现方法。通过数据虚拟化平台,能够将各种异构、异质的数据进行重新整合、抽象,然后定义出新的数据对象供上层使用,有效解决了大数据环境下数据融合的问题。

In order to solve the issues faced by big data, such as diversity of data sources, large scale of data,heterogeneous data, data extraction, data sharing and integrating, a method for data fusion using data virtualization to design and construct data virtualization platform was proposed. Firstly, the design framework of data virtualization platform was given from a global view. Secondly, the scheme was introduced that the heterogeneous data was processed in the source data layer.Thirdly, the structure of the virtual data object layer was analyzed, and the methods of data extracting and writing back were discussed in detail. Finally, the mechanism and method of realizing are expounded in the application interface layer. The problem of data fusion in big data environment was effectively solved through the data virtualization platform, by reintegrating and abstracting the heterogeneous data and then new data objects were defined for upper level using.