The prometheus metric "kubevirt_allocatable_nodes" appears to be tracking arbiter nodes in a Two Node +Arbiter cluster configuration. Arbiter nodes should not count towards kubevirt_allocatable_nodes as arbiter nodes are not designed to run user workloads. As an example, here is the output of a `oc get nodes` for a Two Node cluster:

```
$ oc get nodes
NAME       STATUS   ROLES                         AGE   VERSION
arbiter1   Ready    arbiter                       19h   v1.33.6
master1    Ready    control-plane,master,worker   19h   v1.33.6
master2    Ready    control-plane,master,worker   19h   v1.33.6
```

4.20.3

every time

1. Build a Two Node +Arbiter cluster
2. Install OpenShift Virt on the cluster
3. Using "Observe-Metrics", check the value of "kubevirt_allocatable_nodes" 

The results will show that there are THREE allocatable nodes, however only two of the nodes in the cluster should be used to allocate virtual machines. 

The reported number of allocatable nodes should be reported as TWO.

This was also discussed in the slack thread here: https://redhat-internal.slack.com/archives/C068X44C8VB/p1770733245892159 This bug _may_ also be contributing to false Console alerts "LowVirtAPICount" and "LowVirtControllerCount"