We upgraded a cluster from 4.12 to 4.19 and attempted to run a VM with UDN enabled. The VM networking did not work out-of-the-box.

After investigation, we found the issue was caused by the masquerade subnet size, which had not been explicitly set in the `network/cluster` CR.

By updating the CR as documented here and restarting the VM, egress connectivity (ping) began working as expected.

Proposal:
Add a console alert (or similar notification) when:

• UDN is in use, and
• The configured masquerade subnet size is too small to support UDN.

This would proactively inform administrators of a misconfiguration that currently only surfaces at runtime, reducing troubleshooting time.

Or

Auto updating the masquerade subnet size as part of the OCP upgrade.