Bleeding Edge
Product
Resources
GitHubStart free

AWS Image Builder

Building container images with AWS CodeBuild

The AWS image builder is an image builder flavor provided by the ZenML aws integration that uses AWS CodeBuild to build container images.

When to use it

You should use the AWS image builder if:

How to deploy it

Would you like to skip ahead and deploy a full ZenML cloud stack already, including the AWS image builder? Check out the in-browser stack deployment wizard, or the ZenML AWS Terraform module for a shortcut on how to deploy & register this stack component.

How to use it

To use the AWS image builder, you need:

  • The ZenML aws integration installed. If you haven't done so, run:

    zenml integration install aws

  • An S3 Artifact Store where the build context will be uploaded, so AWS CodeBuild can access it.

  • Recommended: an AWS container registry where the built image will be pushed. The AWS CodeBuild service can also work with other container registries, but explicit authentication must be enabled in this case.

  • An AWS CodeBuild project created in the AWS account and region where you want to build the Docker images, preferably in the same region as the ECR container registry where images will be pushed (if applicable). The CodeBuild project configuration is largely irrelevant, as ZenML will override most of the default settings for each build according to the AWS Docker build guide. Some example default configuration values are:

    • Source Type: Amazon S3

    • Bucket: The same S3 bucket used by the ZenML S3 Artifact Store.

    • S3 folder: any value (e.g. codebuild);

    • Environment Type: Linux Container

    • Environment Image: bentolor/docker-dind-awscli

    • Privileged Mode: false

The user must take care that the Service Role attached to the CodeBuild project also has the necessary permissions to access the S3 bucket to read objects and the ECR container registry to push images (if applicable):

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "s3:GetObject",
                "s3:GetObjectVersion"
            ],
            "Resource": "arn:aws:s3:::<BUCKET_NAME>/*"
        },
        {
            "Effect": "Allow",
            "Action": [
                "ecr:BatchGetImage",
                "ecr:DescribeImages",
                "ecr:BatchCheckLayerAvailability",
                "ecr:GetDownloadUrlForLayer",
                "ecr:InitiateLayerUpload",
                "ecr:UploadLayerPart",
                "ecr:CompleteLayerUpload",
                "ecr:PutImage"
            ],
            "Resource": "arn:aws:ecr:<REGION>:<ACCOUNT_ID>:repository/<REPOSITORY_NAME>"
        },
        {
            "Effect": "Allow",
            "Action": [
                "ecr:GetAuthorizationToken"
            ],
            "Resource": "*"
        },
    ]
}

  • Recommended: Grant ZenML access to trigger AWS CodeBuild builds by registering an AWS Service Connector with the proper credentials and permissions, as covered in the Authentication Methods section. If not provided, the AWS credentials will be inferred from the environment where the pipeline is triggered.

We can register the image builder and use it in our active stack:

zenml image-builder register <IMAGE_BUILDER_NAME> \
    --flavor=aws \
    --code_build_project=<CODEBUILD_PROJECT_NAME>

# Register and activate a stack with the new image builder
zenml stack register <STACK_NAME> -i <IMAGE_BUILDER_NAME> ... --set

You also need to set up authentication required to access the CodeBuild AWS service.

Authentication Methods

Integrating and using an AWS Image Builder in your pipelines is not possible without employing some form of authentication. If you're looking for a quick way to get started locally, you can use the Local Authentication method. However, the recommended way to authenticate to the AWS cloud platform is through an AWS Service Connector. This is particularly useful if you are configuring ZenML stacks that combine the AWS Image Builder with other remote stack components also running in AWS.

This method uses the implicit AWS authentication available in the environment where the ZenML code is running. On your local machine, this is the quickest way to configure an AWS Image Builder. You don't need to supply credentials explicitly when you register the AWS Image Builder, as it leverages the local credentials and configuration that the AWS CLI stores on your local machine. However, you will need to install and set up the AWS CLI on your machine as a prerequisite, as covered in the AWS CLI documentation, before you register the AWS Image Builder.

Stacks using the AWS Image Builder set up with local authentication are not portable across environments. To make ZenML pipelines fully portable, it is recommended to use an AWS Service Connector to authenticate your AWS Image Builder to the AWS cloud platform.

Customizing AWS CodeBuild builds

The AWS Image Builder can be customized to a certain extent by providing additional configuration options when registering the image builder. The following additional attributes can be set:

  • build_image: The Docker image used to build the Docker image. The default is bentolor/docker-dind-awscli, which is a Docker image that includes both Docker-in-Docker and the AWS CLI.

If you are running into Docker Hub rate-limits, it might be a good idea to copy this image to your own container registry and customize the build_image attribute to point to your own image.

  • compute_type: The compute type used for the CodeBuild project. The default is BUILD_GENERAL1_SMALL.

  • custom_env_vars: A dictionary of custom environment variables to be set in the CodeBuild project.

  • implicit_container_registry_auth: A boolean flag that indicates whether to use implicit or explicit authentication when authenticating the AWS CodeBuild build to the target container registry:

    • when this is set to true (default), the builds will be configured to use whatever implicit authentication credentials are already available within the build container. As a special case for ECR registries, the service IAM role attached to the CodeBuild project is used to authenticate to the target ECR container registry and therefore the service role must include the necessary permissions to push images to the target ECR registry.

    • when set to false, the credentials attached to the ZenML Container Registry stack component in the active stack will be set as build environment variables and used to authenticate to the target container registry. This is useful when the target container registry is not an ECR registry or when the service role attached to the CodeBuild project does not have the necessary permissions to push images to the target ECR registry. This works best when the ZenML Container Registry stack component is also linked to the external container registry via a Service Connector.

ZenML Scarf
PreviousKaniko Image BuilderNextGoogle Cloud Image Builder

Last updated