Information Technology Reference
In-Depth Information
Not All Data Sources Are Alike
Some projects, or portions of projects, don't use a database
exactly the way that we define it in this chapter. However, most
projects need to persist data, be it in a flat file, an XML file, a
binary file, or an RDBMS. Regardless of your chosen persistent
store, the principles of CDBI apply.
As a first step in describing how to automate database integration
with CI, we start by describing how to incorporate database integration
into a build process. The scripts used to build, configure, and populate
a database need to be shared with the rest of the project team, so we
discuss which database files are committed to a version control reposi-
tory. Automating a database integration build process solves only part
of the problem, so we go one step further by rebuilding the database
and data at
every
software change—making the verification process
continuous. If a team is adopting CDBI for the first time, most people
on a project will probably need to modify their development practices,
so we finish the chapter looking at effective CDBI practices.
Refactoring Databases
The topics covered in this chapter could even be the subject of a
separate book.
1
Other materials already make the case for treat-
ing your database as just another type of source code that is
managed through the version control repository. This chapter
gives you the essentials to automate and run database integra-
tion processes continuously.
1. In fact, Scott Ambler and Pramod Sadalage have much more in a book called
Refactoring Databases
. Martin Fowler and Pramod Sadalage wrote about similar
topics in “Evolutionary Database Design,” at
www.martinfowler.com/articles/
