The Data Warehouse Toolkit: Practical Techniques for Building Dimensional Data Warehouses

Ralph Kimball's The Data Warehouse Toolkit translates the author's extensive real-world design experience into a truly useful guide to building data warehouses for today's corporations. Written in a no-nonsense prose style, the author gives developers and managers the basics of designing, building, and running data warehouses that are effective and useful to upper management.

This book begins with the fundamentals of what data warehouses are and how they differ from traditional databases used in online transaction processing (OLTP). (Basically, data warehouses look at trends in millions of database records drawn from across organizations.) Once the author establishes this foundation, he solidly introduces the issues of data-warehouse design, using samples drawn from a wide variety of industries, including grocery stores, inventory warehouses, shipping, financial services, cable TV subscriptions, insurance, and travel. The author succinctly describes the relevant "dimensions" of each industry. (These are the values or features that management will most likely want to track for each industry. The data warehouse must be prepared to answer these "questions.")

The last section of the book covers the process of actually designing a data warehouse, from conducting user interviews to planning for the hardware needed to run a data warehouse (where databases easily run in gigabytes and include millions of records). This text also includes a discussion of how to create effective front ends for users (which will report the results of data-warehouse queries).

For its scope and intelligence, The Data Warehouse Toolkit is required reading for any developer or manager who wants to get a perspective on data warehouses before building one. --Richard Dragan