Posted On: Dec 17, 2018

Three new features for AWS ParallelCluster, available today, will help customers further simplify the process of creating and managing HPC clusters on Amazon Web Services (AWS).

AWS Batch Support: Customers can now use AWS Batch as a job scheduler in AWS ParallelCluster and submit distributed computing jobs through AWS ParallelCluster to AWS Batch. AWS Batch enables developers, scientists, and engineers to easily and efficiently run hundreds of thousands of batch computing jobs on AWS. AWS Batch also supports scheduling multi-node parallel jobs, enabling customers to efficiently run single jobs that span multiple Amazon EC2 instances.

Amazon Elastic File System (EFS) integration: AWS ParallelCluster now supports native configuration for EFS to further enhance storage and file management features. Users can create and/or mount EFS directly from AWS ParallelCluster without a post-install script.

RAID Array Configuration Support: AWS ParallelCluster customers can now easily create a RAID 0 or RAID 1 array for use with their HPC cluster. RAID 0 combines multiple drives into a single drive to provide enhanced I/O performance, while RAID 1 combines multiple drives into a single drive for better fault tolerance. RAID arrays can be created from up to 5 Amazon Elastic Block Store (EBS) volumes.

AWS ParallelCluster is available at no additional charge, and you pay only for the AWS resources needed to run your applications. AWS ParallelCluster is released via the Python Package Index (PyPI). AWS ParallelCluster’s source code is hosted under the Amazon Web Services repository on GitHub.  Learn how to launch an HPC cluster using AWS ParallelCluster here.