RJ Atkinson wrote:
On 3 Apr 2009, at 12:49, Wes Beebee (wbeebee) wrote:
Either say that address amplification is forever off the table as a
requirement for IPv6 mechanisms in the space (and I encourage you to try
to get WG consensus on this matter!), define a better mechanism to deal
with address amplification than IPv6-to-IPv6 NAPT, or give up and write
an Internet Draft that defines NAPT66.
Wes,
I have no idea how your quote above relates in any way to anything
that I said. Nothing I said today obviously relates to the topic of
"address amplification".
If I have followed Ran's presentation, 'address amplification' is
orthogonal.
That is provided that address amplification for IPv6 is not engineered
through prefixes longer than /64. The thought truely scares me, based on
where IPv6 addressing has evolved over the last decade.
IMNSHO, there are two solutions to the drivers (at least as I understand
them) for address amplification:
pnat for IPv6 like is done for IPv4.
IPv6 prefixes shorter than /64. Most likely equal to or shorter than /62.
I vote strongly for the later. Anyone that needs address amplification
(and I believe this will apply to all home gamers) should vote with
their pocketbook and not go with ISPs that only allocate /64 or single
/128s. We would have to put this into a BCP to provide the technical
backing for a particular business model.
_______________________________________________
nat66 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nat66
