After much diagnosing we had to conclude the host system was unfortunately not able to recover, the CPU in the system was completely dead. We are sourcing a spare CPU which will be installed, after which we can get the system up. We estimate 30-45 minutes until customer instances are up again.