View a markdown version of this page

Bringing existing resources into Amazon SageMaker Unified Studio - Amazon SageMaker Unified Studio
AWS Glue Data CatalogsAmazon S3 dataAmazon Athena workgroups and saved queriesAmazon EMR on EC2 clustersAmazon EMR Serverless with Trusted Identity PropagationAmazon Redshift clusters and Amazon Redshift Serverless workgroupsAmazon SageMaker AI resourcesAWS IAM roles

Bringing existing resources into Amazon SageMaker Unified Studio

You can bring in existing resources to your Amazon SageMaker Unified Studio project by using the Data and Compute pages in your project, or by using scripts provided in GitHub.

Examples of resources you can bring into Amazon SageMaker Unified Studio are listed below.

AWS Glue Data Catalogs

  • A GitHub script for bringing this resource into Amazon SageMaker Unified Studio can be found here.

Amazon S3 data

To bring your existing Amazon S3 data and use it in Amazon SageMaker Unified Studio, follow the guide in the link below. This guide explains how you can configure permissions and customize role assignments for Amazon SageMaker Unified Studio to access your Amazon S3 data from a project.

  • A GitHub script for bringing this resource into Amazon SageMaker Unified Studio can be found here.

  • To add Amazon S3 connections using the Amazon SageMaker Unified Studio interface, see the Data storage section.

Amazon Athena workgroups and saved queries

  • A GitHub script for bringing this resource into Amazon SageMaker Unified Studio can be found here.

Amazon EMR on EC2 clusters

Amazon EMR Serverless with Trusted Identity Propagation

If you have an existing Amazon EMR Serverless application and want to connect it to your Amazon SageMaker Unified Studio project with Trusted Identity Propagation (TIP) enabled, you must use the AWS CLI. TIP propagates user identity context to Amazon EMR Serverless, enabling fine-grained data access control based on the authenticated user rather than a shared IAM role.

The Amazon SageMaker Unified Studio UI supports connecting existing Amazon EMR Serverless applications, but does not expose the TIP configuration. To create a TIP-enabled connection, use the following procedure.

Before you begin, make sure that you have the following:

To create a TIP-enabled Amazon EMR Serverless connection
  1. Gather your identifiers

    Collect the following values from your environment:

    • Domain ID – In the Amazon SageMaker Unified Studio console, choose Domains, then choose your domain to view the domain details.

    • Project ID – In the Amazon SageMaker Unified Studio console, navigate to your project. The project ID is displayed in the project details.

    • Environment ID – Run the following command:

      aws datazone list-environments \
  --domain-identifier "domain-id" \
  --project-identifier "project-id" \
  --region region

    • Application ARN – In the Amazon EMR console, choose Serverless, then Applications. Alternatively, run:

      aws emr-serverless list-applications
  2. Create the TIP-enabled connection

    Run the following command to create the connection with Trusted Identity Propagation enabled:

    aws datazone create-connection \
  --domain-identifier "domain-id" \
  --environment-identifier "environment-id" \
  --name "connection-name" \
  --enable-trusted-identity-propagation \
  --props '{ "sparkEmrProperties": { "computeArn": "serverless-application-arn"} }' \
  --region region
  3. Verify the connection

    Confirm the connection appears in the Amazon SageMaker Unified Studio UI:

    1. Navigate to your project in Amazon SageMaker Unified Studio.

    2. Choose Compute, then Data processing.

    3. Verify that the new Amazon EMR Serverless connection shows status Active.

    Alternatively, verify through the CLI:

    aws datazone get-connection \
  --domain-identifier "domain-id" \
  --identifier "connection-id" \
  --region region

    Confirm that "enableTrustedIdentityPropagation": true appears in the response.

Amazon Redshift clusters and Amazon Redshift Serverless workgroups

Amazon SageMaker AI resources

  • Amazon SageMaker Studio users can bring their existing SageMaker AI domains, user profiles, and spaces into Amazon SageMaker Unified Studio. This integration also supports additional SageMaker AI resources such as training jobs, machine learning pipelines, models, inference endpoints, and more. A package of GitHub scripts for bringing Amazon SageMaker AI resources into Amazon SageMaker Unified Studio can be found here.

AWS IAM roles

Use the utility script in GitHub to configure permissions and customize role assignments for Amazon SageMaker Unified Studio.

  • A GitHub script for bringing this resource into Amazon SageMaker Unified Studio can be found here.

Note

Review the prerequisites carefully before proceeding to execute the script. Ensure that you save your work and that you do not have any running tasks or processes such as reconfiguring a JupyterLab space or creating a new compute resource. These processes might get interrupted or cause the script to fail.