AWS PCS clusters
An AWS PCS cluster consists of the following components:
-
Managed instances of the HPC system scheduler software, such as the Slurm control daemon (
slurmctld). -
Components that integrate with the HPC system scheduler to provision and manage Amazon EC2 instances.
-
Components that integrate with the HPC system scheduler to transmit logs and metrics to Amazon CloudWatch.
These components run in an account managed by AWS. They work together to manage Amazon EC2 instances in your customer account. AWS PCS provisions elastic network interfaces in your Amazon VPC subnet to provide connectivity from the scheduler software to Amazon EC2 instances (for example, to support scheduling batch jobs on them and enabling users to run scheduler commands to list and manage those jobs).
Topics
