Before configuring an ANSYS RSM Cluster (ARC), it is recommended that you review the important considerations below, and ensure that any prerequisites are met.
ANSYS RSM Cluster (ARC) has basic job scheduling capabilities and provides a simple HPC workflow. However, it is not intended to be a replacement for a commercial scheduler. If you have integrated RSM with a third-party scheduler in the past, or require advanced functionality that the ARC system does not offer, you should continue to work with your IT department to determine which commercial solution will best fit your simulation and business needs. For information about configuring RSM for use with an existing third-party scheduler, see RSM Integration with a Cluster.
The ARC system is not available as a standalone product. It is installed and intended to be used with RSM, meaning that jobs can only be submitted to an ARC via RSM from an ANSYS client application.
RSM must be installed on the machine that you are designating as the submit host, as well all machines that will accept jobs. By default, RSM is automatically installed with ANSYS Workbench products when you use the standard ANSYS product installation.
If creating a multi-node ARC, all execution nodes must be running on the same platform (Windows or Linux).
You cannot run two versions of an ARC (for example, 18.0 and 18.1) at the same time. To ensure that the correct ARC version is used when jobs are submitted to an ARC via RSM, you should set the ARC_ROOT environment variable. Refer to Setting the ARC_ROOT Environment Variable for ANSYS RSM Cluster (ARC) Job Submission.