Research On Management Architecture Of Data Warehouse Metadata

In data warehouse, metadata is used to describe business data structure and building methods, which is a ligament of all data warehouse components. Metadata management is a hot area which needs to pay serious attention on the whole development of the data warehouse, and also needs a strong support of systematic, scientific methods, techniques and tools.This paper analyzes the architecture and data model of data warehouse, and proposes a solution for data warehouse metadata management by researching on the metadata warehousing theory and practice. This paper discusses the following aspects:(1) Figure out the scope, role and management principles of metadata in data warehouse.(2) Summarize the metadata storage model features by analyzing the architecture and the operational data model of data warehouse, and then choose the method of metadata modeling.(3) Discuss the standards of metadata. Propose a federal architecture of metadata management based on specification integrated advantages of centralized and distributed management architecture.(4) Design and implement a data warehouse metadata management system based on the storage and application hierarchy. And propose a solution for metadata storage which supports double-directed migration.This paper focuses on the metadata modeling and the design of storage architecture. Comprehend the connotation of metadata management in data warehouse project from different views respectively. In order to achieve a more robust and versatile metadata management solution, this paper utilizes flexibility methods while observing metadata standard.