Deployment Planning of OpenShift and licensing considerations

Hello everyone,

Thanks for this fantastic community and very much looking forward to contribute to various discussionss.

Have gone through the subscription guide to understand the licensing models to help with planning and optimising some deployments for further growth
https://www.redhat.com/en/resources/self-managed-openshift-sizing-subscription-guide

I understand OCP is licensed by Cores for non bare metal deployments. I understand that Worker nodes are the ones of consideration when it comes to subscriptions and licensing. Are these Cores the Physical Cores that the worker nodes are running on? or the virtual cores presented to the worker nodes VMs ?

The assumption is that for the Worker nodes there could be an elements of overcommitment of CPU resources when a hypervisor such as vsphere is used as the deployment infrastructure.

You could see I've established that that the question focuses on the Worker nodes and not the control plane and infra nodes. Also overcommitment of CPU resources and not of memory.

The Subscription Guide is certainly not clear with explaining this detail.

I'd like to understand the licensing position on this point, the CPU overcommitment of 2:1 or 4:1 is common for some workloads.

I'd think that the various availability levers on the hyper visor need be coordinated with the planned deployment of openshift.

The clear understanding of the licensing approach would help with license cost and hardware cost, which are really required to help with adoption and onboarding of more workloads to the platform.

Looking forward to hearing from you, Thanks