How to Install ARC in a Supercomputing Environment

[jardizzo]

Hi, we convinced our supercomputing facility to give us a special account for running on-prem CI. We found a secure way to do this. We currently have a fixed set of runners executing all CI workflows under the supercomputing account. However, we really need scalable / ephemeral runners to handle the growing demand. Is there a clear set of instructions for installing an ARC in an environment where there is no root access? If not, could we implement our own scalable mechanism using the GH CLI to instantiate ephemeral runners? Any help is greatly appreciated and apologies if this topic has already been addressed.