Partitioning in PostgreSQL was for many years a feature almost entirely ignored, and even today is often thought of only as a means to increase performance on large tables. While it’s certainly useful for this, it’s also an invaluable tool for solving common data maintenance problems. In this article, we’ll examine the feature from this perspective.
Let’s consider a common situation: a table continually ingesting time series data ordered chronologically: purchase orders, forum posts, time-series data, whatever. Most systems contain more than one table like this. And because we live in a world of Big Data, these tables can contain