A single server is a single point of failure, and a single server can only do so much. This final module looks at how PostgreSQL deployments grow: keeping copies of your data on other servers, and spreading the work so the database keeps up as demand rises.
You'll start with replication, where a primary server sends its changes to one or more replicas for resilience and extra read capacity, and see how logical replication lets you copy chosen tables instead of the whole database. Then you'll look at ways of scaling: table partitioning to split very large tables into smaller pieces, and connection pooling to handle many clients without overloading the server.
That completes the lessons. Next, put everything together in the Projects section, where you'll build real things with PostgreSQL.
A tutorial on setting up replication between a primary and a replica server to provide high availability.
What logical replication is, a step-by-step setup with publications and subscriptions, its limitations, how to monitor it, and the best use cases.
Learn which tables need partitioning and how to split a large table into smaller parts to keep queries and maintenance manageable.
A short look at why direct PostgreSQL connections are expensive and how PgBouncer connection pooling helps the server cope with traffic spikes.