Data Warehouse Architecture

Data Warehouse Architecture

Data warehouse architecture is the key factor in building a good data warehouse for your business. Choosing the most suitable data warehouse architecture is a critical task in the data warehouse lifecycle. In this section, we will discuss the most popular data warehouse architectures such as dimensional data warehouse and enterprise data warehouse in a practical approach so you can make a decision to design a good data warehouse architecture for your business.

Popular Data Warehouse Architectures

Enterprise data warehouse architecture -In Bill Inmon’s corporate information factory (CIF) architecture or enterprise data warehouse architecture, the information from various source systems is consolidated into a central repository called an enterprise data warehouse. Data warehouse applications, such as reporting tools, query data from data marts instead of enterprise data warehouses directly.

Dimensional data warehouse architecture – In Ralph Kimball’s data warehouse architecture, data is brought from throughout the enterprise into a central place called a dimensional data warehouse. Like Inmon’s data warehouse architecture, the dimensional data warehouse also has an enterprise focus.  In Kimball’s data warehouse architecture, the data mart is a subset of the tables linking together using star and snowflake schema. Unlike Inmon’s enterprise data warehouse architecture, analytic systems can access data directly from the dimensional data warehouse.

Kimball vs. Inmon in data warehouse architecture – Discusses the differences between Bill Inmon and Ralph Kimball’s data warehouse architectures.

Federated data warehouse architecture – Federated data warehouse architecture provides an effective and practical approach for building a new data warehouse in a heterogeneous environment by integrating legacy data warehouse and business intelligence systems together.

Data Mart – Covers data mart concept and different types of data marts implementations.