Failed even to start migration... any ideas?

Ensure the leader VPN endpoint is reachable by ns7. This is documented here:

https://docs.nethserver.org/projects/ns8/en/latest/system_requirements.html#worker-node-requirements

the API server response contains the leader VPN endpoint: it is a host address with a UDP port number used to set up a Wireguard VPN. The VPN endpoint is configured during the cluster creation procedure. Ensure it is not blocked by other network appliances.

In other words, try to use the leader fqdn ( quasar-2.local.ilogic.gr) not its IP.

I don’t know how much matters in a migration but this looks strange.