caretdev/iris-community-light repository overview

⁠IRIS Light

IRIS Light is a lightweight, minimal InterSystems IRIS® container image designed for use cases where IRIS is required primarily as a high-performance SQL database engine—without the extra components that add overhead, increase image size, and are unnecessary for many modern deployments.

IRIS Light starts from the official vanilla InterSystems IRIS Docker image and removes everything that is not required for core SQL functionality. The result is a significantly smaller, faster-to-deploy, minimal IRIS image perfect for:

• Microservices
• CI/CD pipelines
• Automated testing
• Cloud-native database workloads
• Python / SQL / AI / ML pipelines
• Projects relying on SQLAlchemy-IRIS, JDBC, ODBC, or direct SQL access

IRIS Light is dramatically smaller than the original:

⁠✨ Key Features

• 🚀 Much smaller Docker image size Most optional IRIS components are removed.

• 🧩 Pure SQL-focused IRIS runtime Ideal for lightweight data engines and modern application stacks.

• 🔌 JDBC, ODBC, and SQLAlchemy-IRIS fully supported All client SQL interfaces continue to work as expected.

• 🧱 Based on official IRIS vanilla image Fully compatible with InterSystems licensing, environment variables, and standard tooling.

• 🛠️ Deterministic image build The build process removes components reliably and reproducibly.

⁠🗂️ What’s Removed

IRIS Light strips out components that are often unused in minimal deployments, including:

• Interoperability / Ensemble components (Production engine, Business Services, Adapters, etc.)

• DeepSee / BI stack

• Full Web Stack

  • Management Portal
  • Web Gateway
  • CSP
  • REST framework
  • Web server modules

• Most non-SQL runtime components

• Other optional development libraries and tools

This ensures the resulting container includes only what is needed to run IRIS as a SQL database engine.

⁠🟢 What’s Kept

IRIS Light preserves only the essential pieces:

• Core IRIS database engine

• SQL engine & query processor

• System classes required for storage and globals

• Authentication needed for basic operation

• JDBC & ODBC connectivity

• Compatibility with:

  • SQLAlchemy-IRIS
  • dbt-iris
  • Liquibase-IRIS
  • typeorm-iris
  • Any app using IRIS purely through SQL

Everything else is removed to reduce footprint.

⁠📦 Usage

⁠Pull (once published)

docker pull caretdev/iris-community-light:latest-em

⁠Run

docker run -d \
  --name iris-light \
  -p 1972:1972 \
  caretdev/iris-community-light:latest-em

⁠Connect via JDBC

String url = "jdbc:IRIS://localhost:1972/USER";

⁠Connect via SQLAlchemy-IRIS

from sqlalchemy import create_engine
engine = create_engine("iris://_system:SYS@localhost:1972/USER")

⁠🏗️ Building Your Own IRIS Light Image

IRIS Light can be generated from any official InterSystems IRIS base image, including:

• Enterprise IRIS (containers.intersystems.com/intersystems/iris:...)
• Community Edition (containers.intersystems.com/intersystems/iris-community:...)
• Any other vanilla IRIS image that follows the standard structure

The project includes a convenient build script make.sh that takes:

1. Base IRIS image (vanilla IRIS, Enterprise or Community)
2. Name of the resulting lightweight IRIS image

⁠🛠️ How It Works

IRIS Light uses the official InterSystems IRIS image as the base and applies a reduction script that:

1. Identifies removable components (web, interoperability, DeepSee, others).
2. Deletes binaries, libraries, classes, and system modules not required for SQL.
3. Adjusts permissions and cleans up leftover metadata.
4. Produces a lightweight image with full SQL compatibility.

The project does not modify the IRIS kernel or SQL subsystem, ensuring compatibility with InterSystems’ supported APIs.

⁠🧪 Testing & Validation

IRIS Light has been verified with:

• SQLAlchemy-IRIS
• Basic SQL workloads
• JDBC applications
• Containerized development environments
• Automated CI pipelines

You can extend the included test scripts or integrate IRIS Light into your own tooling.

⁠⚠️ Notes & Limitations

• Management Portal and all web-based functionality are not available.
• Interoperability/Ensemble-based apps will not run.
• DeepSee/Analytics dashboards are removed.
• Licensing works same as upstream IRIS image.

⁠📜 License

This project provides scripts and tooling to modify the official IRIS image. You are responsible for complying with the InterSystems IRIS license agreement and any redistribution rules.