Chapter 45. Data Warehousing with DB2

IN THIS CHAPTER

• Defining the Basic Terms

• Designing a Data Warehouse

• Populating a Data Warehouse

• Accessing the Data Warehouse

• Managing the Data Warehouse

• The Big Picture

• DB2 Data Warehouse Center

• Materialized Query Tables

• General Data Warehouse Guidelines

• DB2-Specific Data Warehousing Guidelines

Data warehousing is not a particularly new idea. The basic idea behind data warehousing is one that has been performed by IT professionals throughout the years : enabling end users to have access to corporate operational data to follow and respond to business trends. You might be tempted, therefore, to shrug off data warehousing as another of the many industry buzzwords that rise and fall every few years. However, doing so would be a mistake.

The true benefit of data warehousing lies not with the conceptual components embodying the data warehouse, but in the combination of these concepts into a single, unified implementation that is novel and worthwhile. Consider the typical IT shop. Data is stored in many locations, in many different formats, and is managed by many different DBMSs from multiple vendors . It is difficult, if not impossible , to access and use data in this environment without a consistent blueprint from which to operate . This blueprint is the data warehouse.

Data warehousing enables an organization to make information available for analytical processing and decision-making. The data warehouse defines the manner in which data

• Is systematically constructed and cleansed (or scrubbed)

• Is transformed into a consistent view

• Is distributed wherever it is needed

• Is made easily accessible

• Is manipulated for optimal access by disparate processes

In this chapter, I provide a basic overview of data warehousing concepts and terms. However, I do not provide comprehensive coverage of all that is implied by data warehousing.

Additionally, I provide useful guidelines for developers who are building data warehouses using DB2. Some of the guidelines are generally applicable to any RDBMS; however, many of them are tailored specifically to DB2 for z/OS.