Chapter 7: Dimensional Modeling
Normalizing data is not always the best strategy when designing a relational database. We already mentioned several times that normalizing data is beneficial for an OLTP workload. OLTP workloads are workloads of primary processes, that is, of line-of-business processes.
Databases normalized to the third normal form turned out to be bad for query performance when we started doing more analytical queries on the data. Dimensional modeling came up as an alternative method for designing database table structures. Dimensional modeling leads to a database design optimized for analytics. For instance, the resulting star schema is the ideal table structure for Power BI.
This chapter is all about dimensional modeling and the resulting star schemas. We will learn about the following topics:
- Background to dimensional modeling
- Steps to get to a star schema database model
- Designing dimension tables
- Designing fact tables
- Using a Kimball...
