In OPNET-133, this PR merged and now if the API or Ingress VIPs are on a subnet that is not part of any interface on a worker node, the worker will be labelled as a remote-worker.

This is problematic in the context of External Load-Balancer where we don't use Keepalived anyway and relocate the VIP management to an external system, therefore the VIP can be part of any subnet, this is not OpenShift's business.

The problematic code is here: https://github.com/openshift/baremetal-runtimecfg/pull/207/files#diff-2b9ef0949e77d903141e49d824606ea166ffe2e95ad18a302212a49a17149191R116-R122