PostgreSQL Client–Server Architecture

Relax. We'll talk through this in plain words — no textbook voice.

What you'll walk away with

Explain the core ideas behind PostgreSQL Client–Server Architecture
Run the sample SQL or command and verify its output
Apply the technique to the Tutorial Platform and production scenarios

Build the mental model

In PostgreSQL, the server process owns the data files and accepts client connections. `psql`, an application server, or a GUI tool acts as a client that sends SQL and receives results. In PostgreSQL terminology, a “cluster” is the collection of databases managed by one server instance, not a Kubernetes cluster.

Connect it to a real scenario

Think of a browser calling an API, then the API server querying PostgreSQL over a database connection. The browser should never receive direct database access. Authentication, validation, and authorization live at the application boundary, with roles and constraints providing another layer inside the database.

The server manages data files
Clients send SQL and receive results
A database port must not be exposed to public browser clients

Try the working example

text

Browser
  └─ HTTPS → Node.js API
                └─ PostgreSQL connection → postgres server
                                             ├─ tutorial_platform
                                             └─ postgres

You should see

You can distinguish clients, servers, connections, clusters, and databases.

5-minute try-it

Draw the browser → API → database data flow for an application you have used.

One important caution

Do not treat a database cluster, database, and schema as synonyms; they have different scopes and permission boundaries.

PostgreSQL — Architectural Fundamentals — PostgreSQL Global Development Group

Easy traps

Do not treat a database cluster, database, and schema as synonyms; they have different scopes and permission boundaries.
Validate sample code on a local or test database with recoverable backups before applying it to production data.

Exercise

Draw the browser → API → database data flow for an application you have used.

You'll know it worked when: You can distinguish clients, servers, connections, clusters, and databases.