I lost approx. 1h this evening with the Error
dhcpcd[139]: eth0: ipv6_start: Cannot assign requested address.
Nothing worked. I tried Managed Host Configuration (checked), Unmanaged Host configuration (unchecked), DHCPv6, SLAAC, Static. Nothing worked.
In desperation and seeing this Post
Logs shows follows:
And then container failed to start.
Make sure you have no same MAC address on your host/CT/VM.
So I stumbled onto the solution using your insight. Turns out that the MAC address used was unique, however it was typed in with lowercase letters. Changing to uppercase did the trick. Something I have never had a problem with before, but apparently it is a problem in this instance.
Thank you for your help!
(in my Case the MAC Address was already all uppercase though)
I just stopped the LXC Container &
cleared the MAC Address for net0. That caused Proxmox VE to automatically generate a new MAC Address.
That did the Trick. After reboot, the error was gone, and I got all 4 IPv6 addresses (ULA, GUA_WAN1, GUA__WAN2, Link-Local) assigned now with DHCPCD

.
Currently the Configuration is still set to "Host-Managed" with Static IPv4 and Static IPv6 (in PVE Configuration Dialog),
/etc/network/interfaces with only IPv4 enabled (as suggested by this Thread) and
touch /etc/network/.pve-ignore.interfaces