Seven Databases in Seven Weeks

Genres Redux

We’ve seen that how databases store their data can be largely divided into five genres: relational, key-value, columnar, document, and graph. Let’s take a moment and recap their differences and see what each style is good for and not so good for—when you’d want to use them and when to avoid them.

Relational

This is the most common classic database pattern. Relational database management systems (RDBMSs) are set-theory-based systems implemented as two-dimensional tables with rows and columns. Relational databases strictly enforce type and are generally numeric, strings, dates, and uninterpreted blobs, but as we saw, PostgreSQL provided extensions such as array or cube.

Good For:

Because of the structured nature of relational databases, ...