No problem at all. Based on this test and the error message above, I guess that there must be some network/firewall settings on that server that are not set properly.
The https://10.90.106.1:8443/ service is supposed to be the lxd daemon listening on the lxdbr0 interface. If your metal installation was implementing a firewall, it could be blocking that. Adding a rule to allow 10.90.106.0/24 (or whatever your lxdbr0 CIDR is) to access 10.90.106.1 tcp port 8443 would possibly help. Or the lxd daemon may have frozen, and you can check the logs of that daemon, or recycle it. Note that this will restart all of your lxd containers that are running.