An ETL Metadata Model for Data Warehousing

Abstract

Metadata is essential for understanding information stored in data warehouses. It helps increase levels of adoption and usage of data warehouse data by knowledge workers and decision makers. A metadata model is important to the implementation of a data warehouse, the lack of a metadata model can lead to quality concerns about the data warehouse. A highly successful data warehouse implementation depends on consistent metadata. This article proposes adoption of an ETL (extracttransform-load) metadata model for the data warehouse that makes subject area refreshes metadata-driven, loads observation timestamps and other useful parameters, and minimizes consumption of database systems resources. The ETL metadata model provides developers with a set of ETL development tools and delivers a user-friendly batch cycle refresh monitoring tool for the production support team.