RE: FW: I-D ACTION:draft-aoun-v6ops-natpt-deprecate-00.txt

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

I agree with Tim below and key point Elwyn made regarding stifling IPv6
deployment.  NAT-PT should be deprecated and SIIT should be left as tool
to build translation when no other choice is possible.  

Lets also be clear about proxies to they are a performance nightmare
worse than NAT and will not solve the E2E security trust problem, when
IPsec is used, unless a decrypt engine is supported on the inbound and
encrypt engine on the outbound.  I can get into more technical details
if we must but I think that should be clear.  Proxies should also not be
elevated as a solution for IPv6 either I detest them as bas as NAT.

But, a use case is as follows:

Premise 1:
Embedded hardware system cannot be upgraded to be IPv6 capable, which
means can support a dual stack.  Reasons could be it uses ASIC IP stack
technology, or the download software or firmware simply cannot add a new
stack to create an IPv6 capable node.

Premise 2:
The network function or applicability to the user provided by the IPv4
legacy system is absolutely required for operations and must be
supported as legacy system.

Premise 3:
The new IPv6 capable systems simply have no more IPv4 addresses to
communicate with the legacy IPv4 system above.

Premise 4:
The new IPv6 capable systems network simply is unable to receive IPv4
packets from any legacy system.

/jim

 

> -----Original Message-----
> From: owner-v6ops@ops.ietf.org 
> [mailto:owner-v6ops@ops.ietf.org] On Behalf Of Tim Chown
> Sent: Sunday, October 10, 2004 5:22 AM
> To: 'V6OPS'
> Subject: Re: FW: I-D ACTION:draft-aoun-v6ops-natpt-deprecate-00.txt
> 
> On Sun, Oct 10, 2004 at 10:16:04AM +0200, Elwyn Davies wrote:
> > 
> >    It  would be useful to see Shivkumar's use cases so that 
> they could be
> >    considered  for  inclusion  in  the  relevant transition 
> analysis.  If
> >    people  are using NAT-PT in a particular way then we 
> need to give them
> >    a workable alternative before ditching NAT-PT.
> 
> I think this is key - the enterprise analysis in v6ops is 
> happening now.
> However,
> 
> 	a) noone is commenting on the work
> 
> 	b) at present, there is no cited case for NAT-PT in the 
> solutions
> 
> I would thus strongly urge those who feel NAT-PT is important 
> to state the
> scenarios and cases where it should be recommended.   It is 
> not at all clear
> what those cases are at present.
> 
> Tim
> 
>