ietf
[Top] [All Lists]

Re: [BEHAVE] [Softwires] Last Call: <draft-ietf-behave-v4v6-bih-06.txt> (Dual Stack Hosts Using "Bump-in-the-Host" (BIH)) to Proposed Standard

2011-09-28 14:14:31
On Wed, Sep 28, 2011 at 11:20 AM, Rajiv Asati (rajiva) 
<rajiva(_at_)cisco(_dot_)com> wrote:
Hi Cameron,

Very interesting (& clever indeed).


How does this clever code ensure that all but a few (pesky apps)
continue to use IPv6 interface instead of the NAT46 interface?

Rajiv,

DNS64 is used.  So anything that can take  a AAAA will use a AAAA and
the native IPv6 path, with or without NAT64 -- as needed.

If the application itself delivers an IPv4 literal via protocols like
MSN or Skype, there is a path and socket made available, that is what
this NAT46 code does.

As i mentioned before, i don't like this.  But, i respect that it
works and it solves a real problem for users of these ipv4-only apps.
I personally find it easy to live with only IP version agnostic apps
that work well in an IPv6-only NAT64/DNS64 network.  I have been
eating this "dog food" for over 18 months.  I am happy to let the
market and eco-system punish apps for not supporting IPv6, and for the
market to reward apps that do support IPv6.

I believe draft-ietf-behave-v4v6-bih-06 has too narrow of a scope to
be useful since it explicitly does NOT support IPv4-only apps talking
to IPv4  servers over an IPv6-only network

Cameron

Cheers,
Rajiv


-----Original Message-----
From: behave-bounces(_at_)ietf(_dot_)org 
[mailto:behave-bounces(_at_)ietf(_dot_)org] On
Behalf Of
Cameron Byrne
Sent: Wednesday, September 28, 2011 2:12 PM
To: Mark Townsley
Cc: Hui Deng; Softwires-wg list; Behave WG; IETF Discussion; Dan Wing
(dwing)
Subject: Re: [BEHAVE] [Softwires] Last Call:
<draft-ietf-behave-v4v6-bih-
06.txt> (Dual Stack Hosts Using "Bump-in-the-Host" (BIH)) to Proposed
Standard

On Wed, Sep 28, 2011 at 9:26 AM, Mark Townsley <mark(_at_)townsley(_dot_)net>
wrote:


+1 ... since the alternative is that apps that require ipv4
sockets and
pass ipv4 literals are stranded on ipv6 only networks.

Running code on the n900 shows that nat464 provides real user and
network benefit

Frankly, I preferred it when you were running IPv6-only without BIH
on your
trial, providing pressure to get rid of all those stranded literals
and
pushing apps to open ipv6 sockets :-/

- Mark

We're still doing that, and IPv6-only is still my philosophical
preference and that is how we are launching the IPv6 + NAT64/DNS64
service into the production mobile network (real soon now).  No change
in that path.

But some "power users" wanted their IPv4-only applications like Skype
to work so they coded a NAT46 work-around for the N900.  It is clever,
it works.

Their process of feeling the pain of a very few pesky IPv4-only apps
and working around it is all documented here:
http://talk.maemo.org/showthread.php?t=60320

Running NAT46 code here: http://code.google.com/p/n900ipv6/wiki/Nat64D

In the end (as well as IPv6-only near term in mobile), IP version
agnostic apps will prove to be more reliable and therefore will get
more market share.

Cameron
_______________________________________________
Behave mailing list
Behave(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/behave

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf

<Prev in Thread] Current Thread [Next in Thread>