Re: Path MTU for draft-ietf-v6ops-mech-v2-00.txt

[Pekka Savola <pekkas@netcore.fi>]

On Sat, 3 May 2003 itojun@iijlab.net wrote:
> >> 	1280 is the only link MTU requirement value appears in RFC2460, and it
> >> 	is the minimum required value for any IPv6 links (which does not
> >> 	require wacky fragmentation header rule at the end of 2460 section 5).
> >> 	i dunno what word other than "simple" would capture this.
> >Correct me if wrong, but it appears the fragmentation header rule must be 
> >included also for the case of MTU 1280 when the IPv4 IP-IP tunnels have DF 
> >bit set.  Which appears to be the typical case, e.g. one major router 
> >vendor doesn't even support anything else.
> 
> 	i don't understand your scenario.  even if there's IPv4-over-IPv4
> 	tunnel with outer DF bit set, ICMPv4 too big won't get propagated
> 	to IPv6-over-IPv4 endpoint, nor IPv6 source.

What I'm trying to say is that there are implementations of IP-IP 
tunneling (subsequently, IPv6-over-IPv4), which always set DF bit in the 
IPv4 header.

In case that IPv4 path MTU is below 1300 bytes, the wacky fragmentation
header rule needs to stay anyway -- unless these implementations are
deemed non-compliant.

So, in practice, PMTUD may be necessary (to lower the MTU) in these cases
even if IPv6 nodes only send packets of 1280 bytes or smaller.

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings