Metadata integration architecture in enterprise data warehouse system

摘要

Due to the increasing complexity of data warehouses, a centralized and declarative management of metadata is essential for data warehouse administration, maintenance and usage. Metadata integration architecture is technical system architecture to achieve metadata management strategy. In this paper, under the background of China Pacific Insurance Company (CPIC) building enterprise data warehouse(EDW) project, we research on metadata integration architecture. First, three kinds of metadata integration architecture were introduced, include: centralized architecture, federal architecture, distributed architecture. Next, applying the federal architecture, we proposed CPIC EDW metadata integration planning. Then again, using CouchDB as the backend, we have developed a metadata management system based on document. As an important enhancements, document based metadata stored in CouchDB in company with metadata stored in MetaStage, constitute the overall CPIC EDW global metadata storage. Finally, we concluded that: ①at this stage, metadata management system in enterprise data warehouse should adopt a federal integration architecture; ② Document based metadata stored in CouchDB can be an important supplementary of the global metadata repository.