We notice that our two instance nodes have various network problems after a live migration to a larger host system. We have currently identified various route problems that are not arriving correctly - in addition, traceroutes and various MTRs are not producing any results apart from ICMP distortions in some cases.
We are continuing to look into the problem.