Database testing – Data Quality
July 2, 2007 Leave a comment
One overlooked area is probably data quality testing.
In fact, from some of the schema designs I’ve seen over the years, I’m sure it is.
Notably in missing primary/foreign/unique keys and other constraints.
To what extent can or should we be testing whether the implementation of a data model enforces the rules, relationships and constraints that it should?
Can any “is”/”has” statements be validated against it?
How should we best go about this?
And are there any tools that can help us do this?
This is definitely needed. And it’s not just one off testing either, because data models evolve constantly and we want to check that a release does not degrade the quality of our data – that
a dropped unique key/index is OK, that a relaxed constraint does not compromise our data.