Ralph Kimball On Reviewing Dimensional Data Models

Ralph Kimball, author of The Data Warehouse Lifecycle Toolkit, has put together a new article on reviewing data warehouse data models. Ralph was an early advocate of dimensional data warehouses, and the modelling techniques detailed in his books and articles are the recommended way that Oracle data warehouses are designed. According to the article;

"People often engage us to conduct dimensional model design reviews. In this column ... we'll provide a laundry list of common design flaws to scout for when performing a review. We encourage you to use this list to critically review your own draft schemas in search of potential improvements."

Some of the areas Ralph looks at in the article include;

• Checking and reviewing the grain of the fact table
• Ensuring only relevant measures and dimension keys are held in the fact table
• Making sure dimension hierarchy information is kept in the dimension and not in the fact table
• Looking at instances of snowflaking and whether this is valid

If you're interested in reading a few more articles by Ralph on dimensional modelling, some of the ones i've found useful include Slowly Changing Dimensions (the classic text on handling changes to dimension values over time), Factless Fact Tables, Fundamental Grains and Declaring The Grain (three articles on fact table design), and Monster Dimensions (how to deal with dimensions with many millions of values). For a general introduction to dimensional modelling and how it works with Oracle data warehouses, check out Ralph's introductory article on dimensional modelling and the OTN Data Warehousing documentation.