When adding new node to Proxmox Cluster, local system might pickup invalid local IP (usually public IP instead of internal, or IPV6 instead of IPv4) and use it in .members file, what results in 595 network problem error messages.
Fix is simple:
- correct /etc/hosts file on affected node
- check with ping that node name returns expected IP address
- $ systemctl restart pve-cluster.service
- systemctl restart corosync.service
- check if /etc/pve/.members is correct