RE: 6to4 vs forwarding IP proto41 in NAT

["Bound, Jim" <jim.bound@hp.com>]

thats one solution. the other is what Jordi proposes to get NAT vendors
to support this outside of the PC and plethora of devices in the home.
/jim

> -----Original Message-----
> From: Christian Huitema [mailto:huitema@windows.microsoft.com] 
> Sent: Tuesday, July 15, 2003 11:23 AM
> To: Tim Chown; v6ops@ops.ietf.org
> Subject: RE: 6to4 vs forwarding IP proto41 in NAT
> 
> 
> On Windows XP, with IPv6 enabled, if you ask to "share a 
> link", the PC will behave as an IPv4 NAT (using ICS) and as a 
> 6to4 router. So I guess there is at least one example of "NAT 
> that is also a 6to4 router."
> 
> ________________________________
> 
> From: owner-v6ops@ops.ietf.org on behalf of Tim Chown
> Sent: Tue 7/15/2003 8:08 AM
> To: v6ops@ops.ietf.org
> Subject: Re: 6to4 vs forwarding IP proto41 in NAT
> 
> 
> 
> The advantage for plain proto41 forwarding is that you don't 
> have to decapsulate for it to work - just pass it on :)  The 
> NAT web config screen can just have a checkbox to forward 
> proto41 and a field to specify the RFC1918 address to forward it to.
> 
> But agreed if the vendor is implementing 6to4, there is no 
> reason why they couldn't choose to do this.
> 
> I wouldn't like to lose the ability to run a tunnel broker 
> client inside my network (for a host or network broker 
> client) because of 6to4 support on the router (which I 
> applaud - and I'd be interested to know which vendors are 
> implementing it... Christian?)
> 
> Of course ideally we want IPv6 natively, but...
> 
> Tim
> 
> On Tue, Jul 15, 2003 at 08:01:45AM -0700, Alain Durand wrote:
> > A NAT box could do the following when receiving an IPv6 packet 
> > encapsulated in IPv4 with proto 41:
> >
> > If IPv6 dst does not belong to the local 6to4 /48 prefix, forward 
> > internally, else decapsulate.
> >
> > Why will will not work?
> >
> >       - Alain.
> >
> 
> 
> 
> 
>