Feelings about NAT64/DNS64?

[Cameron Byrne]

On Mon, Dec 6, 2010 at 8:58 AM, George Bonser <gbonser at seven.com> wrote:
> So far the available solutions for NAT64/DNS64 seem sparse.  It seems to
> me like a good mechanism but so far it looks like it is only available
> on the high-end cisco routers (CGv6), a rather expensive Microsoft
> solution, and an open source solution that is basically a perl script
> and I am unsure will handle any reasonable amount of load.
>
> Anyone investigated this any closer?
>

I am not a vendor and there's all kind's of weird NDAs in the world,
so i won't go into anything specific that i can't point to publicly.

First, i have had NAT64/DNS64 deployed in a mobile network as part of
a beta with about 200 users who have chosen to participate for nearly
a year. NAT64/DNS64 works well in the mobile environment which is
dominated by web and email traffic.  There is a write up on the
experience here
http://tools.ietf.org/html/draft-arkko-ipv6-only-experience-02 .  The
Achilles heel of NAT64 is IPv4 literals, but this problem is not that
big.  I have been trying to get people to document literals they know
of here http://groups.google.com/group/ipv4literals  3 entries so far.

And my trial is here, for anyone interested with a Nokia phone
http://groups.google.com/group/tmoipv6beta

Regarding vendors, expect products across various lines from the big
vendors and small vendors soon. It's the normal mix of vendors that
sell routers and load balancers.  If you ask your account team, they
can brief you on specifics of an individual vendor's plans.  While
there is a perl script, there are also kernel modules from Viagenie in
the freeware camp.

There is some vendor presos here
https://sites.google.com/site/ipv6implementors/2010/agenda


Cameron

> George
>
>