OpenWrt Forum Archive

Topic: wan6 ... why?

The content of this topic has been archived on 4 May 2018. There are no obvious gaps in this topic, but there may still be some posts missing at the end.

So as I dive back into OpenWRT I'm curious as to why we have a wan6 interface. Is it solely so we can specify a different protocol for address assignment? In otherwords:

option proto 'dhcp'

and

option proto 'dhcpv6'

I see in the IPv6 wiki post that if I had both a static IPv4 and a static IPv6 assignment I could do it all in 1 interface. If this truly is the reason it sounds like we need a proto<4> and a proto6 option so we can use a single interface to deal with the different protocols instead of having extra interfaces to confuse muggles like myself.

Thanks in advance.

Its because of netifd architecture.
One logical (netifd) interface supports only one way of external configuration.
Internally netifd is very limited in configuration possibilities and it relies on external scripts in /lib/netifd/proto.
One script for one protocol, one protocol for one interface. Need another protocol - use another logical interface for the same physical device

Thanks for the reply bolvan. Based on your reply I could go either way; a virtual interface per protocol or an interface with protocol independent configurations (like a cisco router).

I wonder if that could be placed in the IPv6 wiki for those of us coming from different architectures? I know I was confused by it.

Thanks!

The discussion might have continued from here.