Understanding the Behavior of Keys
Unique identifiers (a.k.a. “keys”) play an important role in managing data in relational databases; without the appropriate keys, data cannot be reliably added, accessed, updated, or deleted.
Yet, many of the people responsible (data stewards, application managers, etc.) for the data and/or the applications that manipulate data remain oblivious of the subtleties in key behavior. This state of affairs is especially true in cases where data are acquired from an external source, or when data from multiple sources are integrated.
This two-part series of articles aims to provide a set of questions that will help data and application managers better understand how the keys behave.
Michael Scofield begins a two-part series which offers guidelines for analyzing and assessing the quality of keys in databases and files.