Background to Data Vault modeling
We have already discussed nomalizing data and dimensional modeling as possible ways to design relational databases. Normalizing is a great strategy for databases that support line-of-business applications. Dimensionally modeled databases are optimized for reporting. In this chapter, we will look at a third alternative – Data Vault modeling. It optimizes the model for the flexible long-term storage of historical data.
Have a look at Figure 9.1, which is a copy of Figure 7.2:
Figure 9.1 – Classic data warehouse architecture
This figure shows how Ralph Kimball envisioned the data warehouse. The data warehouse is a collection of all the star schemas describing an organization. It has a star for Sales, a star for Marketing, a star for Human Resources, and so forth. It is loaded from the data that originates from line-of-business applications. Whenever two processes use the same dimension, you create a conformed...