Deployment of AWS cluster with custom odyhpc AMIs   (i – preliminary steps)

AWS has developed AWS ParallelCluster, which renders launching HPC clusters in the cloud significantly easier than it was a few years ago. If you are already familiar with AWS ParallelCluster, you can skip this step. Before proceeding, make sure that you have available a copy of the following data: the name of the region where the cluster will be launched along with one of the keypairs available in that region, the AMI ID for that region, and the IAM credentials (aws_access_key_id and aws_secret_access_key).

  1. AWS ParallelCluster Installation – AWS ParallelCluster can be installed either in virtualized or non-virtualized environments. It runs on Linux, Windows or macOS. Although the AWS ParallelCluster team recommends using a virtualized environment, we have installed AWS ParallelCluster in non-virtualized environments and found no problem. Whichever are your choices for environment and OS, your system must have python and pip (or python3 and pip3) installed. If that is not the situation, you can follow specific instructions for any combination of environment and OS available at https://docs.aws.amazon.com/parallelcluster/latest/ug/install.html. Once pip is installed, the command $pip/pip3 will install aws-parallelcluster. The system will show the progress of the installation and list the installed apps at the end of it.

Close Menu