when setting spec.k0s.config.spec.network.dualStack.IPv6serviceCIDR to a prefix >108 and then running k0sctl apply, the k0scontroller service crashes with

 --service-cluster-ip-range[1] is too large; for 128-bit addresses, the mask must be >= 108

but because k0scontroller is not running, running k0sctl apply again results in

 - [ssh] x.x.x.x:22: k0s doesn't appear to be running but has been installed as a service at /etc/systemd/system/k0scontroller.service - please remove it or start the service 

To fix it you have to correct the rendered /etc/k0s/k0s.yaml on the host(s). Not a big deal, but it would be nice to validate the config before applying to avoid cluster downtime