Each node on the cloud has a limit for the number of egress IPs that can be assigned to it.

e.g AWS cluster:

 [{"interface":"eni-00b25ff11a7011abe","ifaddr":{"ipv4":"10.0.160.0/20"},"capacity":{"ipv4":14,"ipv6":15}}] 

We should eventually warn the user when we are at 80% of the allowed limit for a node. We first need a metric to expose this information.

For non-cloud environments, put the remaining amount at max value for float64.

DoD: PR merged downstream in OVN-Kubernetes and SDN and validated by QE for at least one of the cloud providers.

Follow-on: once DoD is met, create a story for an alert.