PostGIS Cookbook - Second Edition

Geospatial sharding

Working with large datasets can be challenging for the database engine, especially when they are stored in a single table or in a single database. PostgreSQL offers an option to split the data into several external databases, with smaller tables, that work logically as one. Sharding allows distributing the load of storage and processing of a large dataset so that the impact of large local tables is reduced.

One of the most important issues to make it work is the definition of a function to classify and evenly distribute the data. Given that this function can be a geographical property, sharding can be applied to geospatial data.