Optimizing Enterprise Data Warehouse Design
Utilizing Dimensional Normal Form

by Michael Schmitz

Description

Dimensional Normal Form is a new approach to Data Warehouse Data Architecture which combines the strengths of both the normalized and the dimensional design paradigms to provide usable, flexible, scalable, and high performing schemas for the Enterprise Data Warehouse. Dimensional Normal Form allows an Enterprise Data Warehouse to be constructed a Data Mart at a time without requiring an intermediate Data Warehouse. While this sounds like the Kimball approach it is differentiated by its completely normalized ETL dimensions and its methods for tracking dimension history and differentiating contextual from detail or audit history.
This seminar fully covers design techniques for Data Warehousing and BI solutions based on the Dimensional Normal Form Data Architecture and discusses the pros and cons of the many design decisions that must be made. History considerations are discussed in detail along with their impact on schema design.
Various schema design examples are presented and discussed. The seminar participants will also be given design exercises will have their solutions analyzed and discussed by the group.
The seminar will also present contrasting Data Warehouse architectures and discuss the variations in physical design that are required for different Data Warehousing environments.
The following seminar on “Template-Driven ETL” will show how to build, populate and maintain the schemas built following the Dimensional Normal Form Data Architecture.