View a markdown version of this page

Explore Image Builder sample projects on GitHub - EC2 Image Builder

Explore Image Builder sample projects on GitHub

The EC2 Image Builder team maintains the Image Builder samples repository on GitHub. The repository contains complete, deployable sample projects that show how to implement common use cases with infrastructure as code. Samples are available as AWS CloudFormation (CloudFormation) templates, AWS Cloud Development Kit (AWS CDK) applications, Terraform modules, and AWS Task Orchestrator and Executor (AWSTOE) component documents.

Each sample is self-contained, with its own README that covers prerequisites, deployment steps, how the sample works, testing, and cleanup. The tutorials in this chapter show you the image building process from the console. The sample projects show you the same concepts expressed as code that you can deploy, inspect, and adapt.

The repository includes samples for the following areas:

  • Getting started – The smallest complete image pipeline, provided as both a CloudFormation template and an AWS CDK application in the quick-start sample.

  • Use-case kits – Complete pipelines for common scenarios, such as a golden AMI pipeline with scheduled patching and Auto Scaling group refresh. Other kits cover cross-account AMI distribution, private VPC builds, lifecycle cleanup policies, and container base image pipelines that publish to Amazon Elastic Container Registry.

  • Image workflows – Custom workflow samples that add human approval gates to the build and distribution stages, and integrate AWS Step Functions for out-of-band AMI validation.

  • Platform coverage – Golden image pipelines for Windows Server, macOS, and Linux.

  • Components and debugging – AWSTOE component patterns such as reboot-and-resume and build-time secrets. Also includes a script that tests component documents locally. A debugging walkthrough diagnoses a build that fails on purpose.

For the full list of samples, see the index in the repository README on GitHub. The samples are licensed under MIT-0, and the team reviews issues and pull requests on a best-effort basis.