Chapter 5. How schemas grow
This chapter is recommended for
| Business analysts | |
 |
Data architects |
|
Enterprise architects |
| Application developers |
In chapter 4, you learned how to write a Mondrian schema that contains a simple cube. Even a basic cube can support countless analyses; each analysis answers some questions and raises new ones. If that first cube is successful, your business users will come back and ask for more dimensions, more attributes, and more powerful ways to model and analyze the data.
This chapter is about how dimensional models tend to evolve in the real world. Much of that evolution is “more of the same”: adding cubes, dimensions, and measures. A small addition to the model can allow a significant new area of the business to be analyzed. For example, a business user might ask you to add a Referrer dimension to the Sales cube so that they can analyze the effectiveness of social media campaigns.
Mondrian has features that keep schemas manageable as they grow. We’ll look at how you can create and use shared dimensions, and how you can use measure groups to build a cube based on more than one fact table.