Resizing the HPC Cluster

For HPC clusters created in project spaces that support Ansys 2024 R1 and earlier releases

To better meet the computational requirements of a specific problem, you can increase or decrease the number of nodes in an HPC cluster as needed.

Locate the HPC cluster tile in your project space and follow the steps in Resizing an HPC Cluster on Demand in the User's Guide.

Before submitting a job to the cluster, ensure that all nodes are in the Running state and are Ready for scheduling.

To check that the cluster is fully defined and ready, type sinfo from a command prompt on the Controller virtual machine.

If the cluster is ready, the number of nodes reported by sinfo is equal to what you requested, and their state is 'idle'. If the cluster is not ready, wait a few minutes and then try sinfo again. If there is a delay, typically it is only 1 to 2 minutes.

For more details on testing the readiness of a Slurm cluster, see Testing a Slurm HPC Cluster.