No InternalDNS and Hostname in node status for k8s 1.29

[jayesh-srivastava]

/kind bug

What steps did you take and what happened: Have observed inconsistent behaviors in CAPG clusters using Kubernetes version 1.28 or below v/s CAPG clusters using Kubernetes version 1.29 with Cloud Controller Manager. The difference appears in Node.status.addresses where Kubernetes 1.28 or below clusters have all InternalIP, InternalDNS and Hostname as opposed to Kubernetes 1.29 clusters who only have InternalIP. This might create issues for downstream users.

What did you expect to happen: Expected the Node.status.addresses to have all three types of addresses.

Anything else you would like to add: Snippet from cluster using K8s 1.28:

Snippet from cluster using K8s 1.29:

Environment:

• Cluster-api version:
• Minikube/KIND version:
• Kubernetes version: (use kubectl version):
• OS (e.g. from /etc/os-release):

[jayesh-srivastava]

Have opened similar issue at https://github.com/kubernetes/cloud-provider-gcp/issues/773