Effective Methods for Software Testing, Third Edition

Chapter 21. Testing a Data Warehouse

A data warehouse is a central repository of data made available to users. The centralized storage of data provides significant processing advantages but at the same time raises concerns of the data’s security, accessibility, and integrity. This chapter focuses on where testing would be most effective in determining the risks associated with those concerns.

Overview

This testing process lists the more common concerns associated with the data warehouse concept. It also explains the more common activities performed as part of a data warehouse. Testing begins by determining the appropriateness of those concerns to the data warehouse process under test. If appropriate, the severity of the concerns must be determined. This is accomplished by relating those high-severity concerns to the data warehouse activity controls. If in place and working, the controls should minimize the concerns.

Concerns

The following are the concerns most commonly associated with a data warehouse:

• Inadequate assignment of responsibilities. There is inappropriate segregation of duties or failure to recognize placement of responsibility.

• Inaccurate or incomplete data in a data warehouse. The integrity of data entered in the data warehouse is lost because of inadvertent or intentional acts.

• Losing an update to a single data item. One or more updates to a single data item can be lost because of inadequate concurrent update procedures.

• Inadequate audit trail to reconstruct transactions. ...