AWS Parallel Computing Service now supports instance settings for compute node groups
AWS Parallel Computing Service (AWS PCS) now supports instance settings, which let you configure common Amazon EC2 settings for compute nodes directly on a compute node group. This simplifies cluster setup: instead of first creating a launch template in Amazon EC2, you can choose the AMI, security groups, and SSH key pair alongside the rest of your compute node group configuration. You can use the AWS PCS console, AWS CLI, SDK, or infrastructure as code (IaC) tools, and AWS PCS can catch a missing or invalid setting when you make the request.
You control these settings from first setup through later updates. For example, you can pair instance settings with node lifecycle actions that mount shared storage. Later, to patch the AMI or rotate an SSH key, you update the compute node group, and AWS PCS can replace compute nodes after their running jobs finish, eliminating the need to manually drain nodes.
AWS PCS is a managed service that simplifies running and scaling high performance computing (HPC) workloads on AWS using Slurm. You can build complete, elastic environments that integrate compute, storage, networking, and visualization tools, and the service manages cluster updates and provides built-in observability.
Instance settings are available in all AWS Regions where AWS PCS is available. To learn more, see Configuring instance settings in the AWS PCS User Guide.