AWS ParallelCluster 3.2 File System Enhancements and Other Requested Features

AWS ParallelCluster 3.2 release is generally available.

AWS ParallelCluster is a supported and maintained open source cluster management tool that makes it for scientists, researchers, and IT administrators to deploy and manage HPC clusters in the AWS cloud. HPC clusters are collections of tightly coupled compute, storage, and networking resources that enable customers to run large scale scientific and engineering workloads.

Feature enhancements in AWS ParallelCluster 3.2 include:

Multiple file systems mounts. Customers can mount up to 20 Amazon FSx and 20 Amazon EFS existing file systems to a single cluster.
Support for additional Amazon FSx file systems. Customers can mount existing Amazon FSx for OpenZFS, Amazon FSx for NetApp ONTAP, and Amazon FSx for Lustre Persistent_2 file systems to cluster. The ParallelCluster cluster configuration file provides additional sections to configure these file systems.
Memory-aware job scheduling with Slurm. Customers can leverage Slurm’s memory-based scheduling functionality to specify memory requirements on job submission, and control placement of compute instances that match those memory requirements.
Extending flexibility to dynamically update clusters. Customers can dynamically update job queue properties such as AMI without having to stop and start their clusters. For example, through a simple CLI command, user can hot swap an AMI while controlling the cycling of compute nodes.