Mastering Data Warehouse Aggregates: Solutions for Star Schema Performance

Chapter 7. Aggregates and Your Project

The design and implementation of aggregate tables always take place in the context of a project. The project may involve the development of a new warehouse subject area, orbe exclusively focused on adding aggregates to an existing schema. Regardless of project nature, aggregates implementation requires a standard set of tasks and deliverables, which mirror those required by the base schema.

This chapter begins with a look at the process of data warehouse implementation. A dimensional data warehouse is most often implemented in parts known as data marts. If organized around a planned framework of conformed dimensions, this incremental approach avoids the dangers of enterprise scope at one extreme, and a departmental focus on the other. It also allows the introduction of aggregates to be delayed until after the first data mart has been constructed, although this is not required.

Next, the specific implementation tasks and deliverables surrounding aggregates are presented in the context of a data mart development project. These tasks are separated into project stages—strategy, design, build, and deployment—but can easily be reorganized to fit any of the popular development methodologies.

Last, this chapter looks at the ongoing maintenance requirements for aggregate tables within a data warehouse, and considers the change management process by which they are modified.

Much of this chapter's discussion of data warehouse implementation hasbeen ...