Thank you for your replys.

We restarted the mgmt-services via iLo (I assume this does the same as services.sh start/stop). This took about 20 minutes, but worked. Now after migrating the VMs to another host and reboot of the host, everything seems to work fine again.