Amazon SageMaker HyperPod now supports AMI-based node lifecycle configuration for Slurm clusters

Amazon SageMaker HyperPod Updates

Amazon SageMaker HyperPod now supports AMI-based configuration for Slurm cluster nodes, streamlining the setup for AI/ML training workloads. This new configuration method eliminates the need for downloading, configuring, or uploading lifecycle configuration scripts to Amazon S3, significantly reducing cluster creation time.

AMI-based configuration includes essential software like Docker, Enroot, and Pyxis, and configurations such as Slurm accounting, SSH key generation, Slurm log rotation, and user home directory setup. To enable this configuration, omit the LifeCycleConfig block when creating clusters via the CreateCluster API or select "None" under Lifecycle scripts in the SageMaker AI console.

For additional customization, an extension script can be provided, allowing for user configuration, observability, or LDAP integration. This can be configured through the API by specifying the OnInitComplete parameter and SourceS3Uri in the LifeCycleConfig block, or via the console by providing the S3 URI to the extension script in the "Extension script file in S3" field in Custom setup.

This feature is available in all AWS Regions where SageMaker HyperPod is available. For more information, refer to the SageMaker AI developer guide.

Source: AWS release notes

If you need further guidance on AWS, our experts are available at AWS@westloop.io. You may also reach us by submitting the Contact Us form.