Re: "RFC 2461bis" issue: MTU handling

[itojun@itojun.org (Jun-ichiro itojun Hagino)]

> > 	the assumption made in RFC2461 is that the link MTU is constant
> > 	over the link, i guess.  i don't think it is necessary to make
> > 	MTU negotiable between peers, it would complicate too many things.
> What would it complicate?
> Obviously nobody is going to force anyone to implement such an option; 
> it would only be used between consenting hosts.

	you need to have a mechanism if the peer is capable of handling the
	option or not (since it is optional, if i read your comment correctly).
	then, neighbor cache management has to be changed to handle per-peer
	MTU value.

itojun

--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------