Getting Started with Edge Microvisor Toolkit: Repository Overview & Download Guide

[biapalmeiro]

EMT is a powerful, open-source toolkit designed to build immutable (or mutable) Linux OS images tailored to edge workloads — all while leveraging the performance and feature set of Intel® hardware.

In this discussion, we’ll walk through:

• EMT key features
• How the GitHub repo is structured
• How to clone and build your own image

---

What are the key features?

EMT is a downstream of Azure Linux that empowers developers to build rpm-based images using declarative JSON configuration files. Whether you're deploying to containers, bare metal, or Kubernetes, EMT gives you a consistent, high-performance platform that emphasizes security, scalability, and manageability.

Key features include:

• Atomic A/B updates with rollback support

• Immutable image design

• Real-time image support

• dm-verity and Secure Boot

• Flexible build system and custom package support

---

Repository Structure: What’s Inside?

Here's a quick breakdown of the key directories in the GitHub repo:

edge-microvisor-toolkit/
├── SPECS/                 # RPM spec files for packages
├── SPECS-EXTENDED/        # Additional and optional specs
├── SPECS-SIGNED/          # Signed packages for production
├── imageconfigs/          # JSON config files defining image layouts
├── packagelists/          # JSON files listing rpms to include in images
├── toolkit/               # Build system, scripts, and cgmanifest
│   └── scripts/           # Tools for automation and metadata generation
└── docs/                  # Usage instructions and architecture overviews

• SPECS/: Define the behavior and structure of RPM packages used in image builds.

• imageconfigs/: These JSON files describe partition layouts, bootloaders, and packagelist references.

• packagelists/: Modular lists of packages grouped by purpose (e.g., virtualization, GPU support).

• toolkit/: The actual build system that compiles and packages the images.

---

How to Download and Build EMT

You can either download prebuilt images (via the releases page) or build a custom image from source. Here’s how to get started with the build system.

Prerequisites

Make sure you're on Ubuntu 22.04+ and have the following tools installed:

sudo apt update
sudo apt install git make python3-pip rpm build-essential

Clone the Repository

It’s recommended to use a stable release tag instead of main or latest:

git clone --branch <stable_tag_name> https://github.com/open-edge-platform/edge-microvisor-toolkit.git
cd edge-microvisor-toolkit/toolkit

Build the Toolchain

sudo make toolchain REBUILD_TOOLS=y

Build an ISO Image

Choose your image configuration (e.g., edge-image.json, edge-image-rt.json) from the imageconfigs/ folder:

sudo make iso -j8 CONFIG_FILE=./imageconfigs/edge-image.json

The ISO will be generated in the build/ directory and can be flashed to a USB drive or booted in a VM.

---

Customization & Packages

Want to include your own tools, AI inference packages, or drivers? EMT allows you to:

• Create custom packagelists

• Add new RPM SPEC files

• Use dm-verity, FDE, or Secure Boot

• Customize bootloaders and system services

This makes it a powerful base for developers working on ML/AI workloads, industrial control systems, and IoT infrastructure at the edge.

---

Ready to dive in?

📖 Check the Developer Guide
💬 Have feedback or questions? Drop us a comment.