Strange HA behaviour (tested from 2.2 to 3.0)

Hi,

I just discovered a strange behaviour on a two node proxmox HA cluster. Maybe some of you could point me in the right direction.

When I live migrate a running VM from node A to node B, the migration runs without a problem. The VM is running on node B now and I can access the console, the VM runs without a problem. I can ping from the VM to node B, but not any other machine like the default gateway.

The strange thing is: I just have to wait for 5 to 10 minutes and suddenly ping works to all hosts again and the VM is accessible by IP.

This also does not happen when I migrate von node B to node A.

If I'd like to guess, I'd assume some arp related problem at switch level. But this is really just guessing.

Any hints?
Marco