🚀 Pglayers: Composable PostgreSQL Extensions via Docker Layers

Pglayers transforms how you handle PostgreSQL extensions. Instead of wrestling with complex build environments, it provides a system where extensions are treated as stackable Docker layers.

Check it out on GitHub: pglayers/pglayers

---

🛠️ The Problem & The Solution

The official PostgreSQL Docker images are lean, meaning they ship without any extensions. Traditionally, if you needed pgvector, PostGIS, or pg_cron, you were forced to:

• Manually install system dependencies
• Compile source code from scratch
• Rely on bloated third-party images with rigid, pre-defined extension sets

Pglayers solves this by offering a ready-to-use distribution and a toolkit to assemble your own custom image. It sits directly on top of the official postgres images, removing the need for you to manage compilers or build logic.

How it Works: The Layering Concept

Each extension is packaged as a lightweight Docker image containing only the necessary binaries. You simply "stack" these layers using the COPY --from instruction.

Mathematically, the final image $\text{Im}_{final}$ can be represented as the sum of the base image and the selected extension layers: $\text{Im}_{final} = \text{Im}_{postgres} + \sum_{i=1}^{n} \text{Layer}_{ext\_i}$

---

🚀 Quick Start Guide

Depending on your needs, you can choose between pre-configured bundles or a bespoke setup.

Option 1: Ready-to-Use Images

These images come with shared_preload_libraries already configured for convenience.

• Full Suite: Includes all 53 available extensions.
• Azure Profile: Compatible with Azure Database for PostgreSQL (28 extensions).

Deployment Commands:

# Launch the full version (PG 17)
docker run -d -e POSTGRES_PASSWORD=secret ghcr.io/pglayers/pglayers-full:17

# Launch the Azure-compatible version (PG 17)
docker run -d -e POSTGRES_PASSWORD=secret ghcr.io/pglayers/pglayers-azure:17

[!IMPORTANT] Note on Vendor Profiles: Profiles like azure are best-effort approximations intended for local development. They are not exact replicas of managed services; differences in configuration defaults, versions, and platform behavior may exist. Use them for testing, not for mirroring production perfectly.

Option 2: Custom Composition

If you want a lean image, pick only the extensions you actually need.

Example Dockerfile:

FROM postgres:17

# Stack only the required extensions
COPY --from=ghcr.io/pglayers/pgx-pgvector:17 / /
COPY --from=ghcr.io/pglayers/pgx-pg_cron:17 / /
COPY --from=ghcr.io/pglayers/pgx-postgis:17 / /

Build and Execute:

1. docker build -t my-postgres .
2. docker run -d -e POSTGRES_PASSWORD=secret my-postgres

Since the layers are pre-built, no compilation occurs during the build process.

---

📊 Compatibility & Support

(Example representation of the containerization approach)

PostgreSQL Version Status

---

📚 Available Extension Catalog

Below is the comprehensive list of extensions provided by Pglayers.