Add OCP-CNV patching demo

[l3acon]

This should at least be close to ready to use.

[l3acon]

To do's

• take out machine credential
• check out if RHEL VM entitlements can be automatic
• add wait_for to end of provision of VM
  This is more complex than originally thought. We rely on the inventory plugin to set ansible_host, thus we must run the inventory sync first to determine a host to check against SSH being up for. My simple solution to this is simply run the inventory sync twice. ¯_(ツ)_/¯
• fix cnv-inventory credential type to have descriptions the same as OCP type
• timing of VM creation & post CNV operator install, chris ran into an issue where the automation tried to create the VM but the images hadn't pulled yet, so they became stale and never provisioned

[l3acon]

I have found solutions for both of the timing issues:

• For validating operator install, we can repeatedly try to provision a VM, when it finally succeeds the operator is finished initializing and ready.
• For the inventory sync issue, there is a module, kubevirt.vm_info that can be used to wait for a VM to be in the running state.

Additionally, I discussed with both OCP and RHEL SMEs about subscribing RHEL VMs running on OCP-V. There does not seem to be a more elegant method than using activation keys (or a manifest of some kind).

After what feels like a lifetime ago, I believe this is finally ready for review.

[willtome]

LGTM. Do we need to wait for a new EE build from @cloin before merging?

[l3acon]

@IPvSean already merged ee-builds for us, however it seems the github action did not succeed from Sean's merge https://github.com/cloin/ee-builds/actions/runs/9454222805/job/26041308031#step:7:22 and that is causing the inventory bug @willtome discovered.

[l3acon]

The new image built. I am going to do a clean slate run through and merge if there are no issues.