What happened:
When spawning VMs in an IPv6 only Kubernetes cluster using Passt as a network binding plugin.
Everything works as expected, but after ~24 hours the VMs lose the default gateway.

This behaviour was reproduced both on Flatcar Container Linux and Talos (as Guest OS).

What you expected to happen:
The default gateway should not vanish and network connectivity should persist.

How to reproduce it (as minimally and precisely as possible):
Spawn a VM with Passt network plugin. Wait at least 24-36 hours and test network connectivity.
Do not create an alternative static route for ::/0.

Additional context:
On Talos nodes running talosctl get routes the node contains the following route in the beginning:

NODE                          NAMESPACE   TYPE          ID                                                                      VERSION   DESTINATION                                   GATEWAY                     LINK         METRIC
fd01:cafe::5054:ff:fe10:9ad   network     RouteStatus   eth0/inet6/fe80::ecee:eeff:feee:eeee//1024                              1                                                       fe80::ecee:eeff:feee:eeee   eth0         1024

The route will disappear. However the problem can be worked around by adding a static route for ::/0 for eth0.

Environment:

• KubeVirt version (use virtctl version): v1.4.0-alpha.0
• Kubernetes version (use kubectl version): v1.31.1
• VM or VMI specifications: N/A
• Cloud provider or hardware configuration: Bare-Metal
• OS (e.g. from /etc/os-release): Talos 1.8
• Kernel (e.g. uname -a): N/A
• Install tools: N/A
• Others: N/A