Re: [Softwires] Thoughts on new I-D?: Gateway Initiated Dual-stack lite Deployment

[Christian Vogt <christian.vogt at ericsson.com>]

Frank and Sri -

I think you should add a motivation for your proposal to the document.
This motivation should answer the questions:  Why does the NAT need to
be de-coupled from the gateway that terminates the existing tunnel?  Why
not co-locate the NAT with the gateway?  This would explain why you want
to extend the existing tunnel towards the NAT using your proposal.

I see one motivation, which applies to networks with multiple gateways:
In those networks, by de-coupling the NAT from the gateways, the global
IPv4 addresses on the external side of the NAT can be shared more
efficiently compared to co-locating a separate NAT with every gateway.

Are there other motivations for your proposal?

- Christian



On Oct 15, 2009, Frank Brockners wrote:

> FYI. We just wanted to draw your attention to a new draft "Gateway
> Initiated Dual-stack lite" - which outlines a modified deployment
> approach of the  original dual-stack lite. The approach applies to  
> those
> network architectures which already employ some form of tunneling
> technology in the access network. The key idea is to leverage the
> existing access tunnel (which means that architectural changes within
> the access network are limited to the gateway terminating access  
> tunnel)
> and stitch it together with a softwire connecting to the CGN. While
> largely targeted at mobility architectures, the solution could be
> applied to broadband environments as well.
>
> A copy of the draft is available at:
> http://www.ietf.org/id/draft-gundavelli-softwire-gateway-init-ds-lite-00
> .txt
>
>
> We appreciate your thoughts and comments.
>
> Thanks & regards, Sri & Frank