Starring Sakila: Data Warehousing Explained, Illustrated, and Subtitled

MySQL is increasingly being adopted for Business Intelligence (BI) and data warehousing (DWH) purposes. Although the technology as such is far from new, there is still quite noticeable confusion and misunderstanding among MySQL Developers and DBAs surrounding key BI/DWH concepts such as denormalization, dimensional modeling, ETL, and OLAP. Even if the concepts are clear, it is still not clear what the purpose is of a datawarehouse, and what problems it can solve.

The Sakila database is a simple, publicly available sample database of a DVD rental business, distributed by Sun (formerly MySQL AB). It is quite popular for illustrating traditional database schema design issues and SQL queries.

The sakila database makes for a great mini-tutorial on how to design, create, load and use a data warehouse. In 45 minutes, attendees can witness exactly how this is done, and how the result differs from the original sakila schema.

The following outline explains the structure of this talk in detail

• The BI problem: Creating information out of data. Brief introduction explaining information needs of managers and up.
• How data warehousing helps to solve the BI problem.
• Anatomy of a data warehouse. Introduction to dimensional modeling, differences with operational database design.
• Major physical implementations of data warehouses: star schemas v.s. snowflakes
• A star is born: dimensional models for rentals and payments from the sakila database
• Filling the data warehouse: extraction, transformation and loading
• The result: making managers happy with dynamic pivot tables, slice and dice and drilldown.