Re: [dhcwg] Re: [Int-area] DCHP-based authentication for DSL?

[Richard Pruss <ric at cisco.com>]

Hi Ted,

The message you quote seems to be a what about the clients that do not 
support a new authentication method.  The answer is they work as always 
with PPPoE.

The requirement to remove PPPoE from the DSL architecture is driven by 
part technical and part emotional reasoning.  On the technical side you 
see a real need to remove the tunnelling aspect of PPPoE so that 
services (like multicast) can be injected in the layer 2 network.  The 
new services are supported by new equipment and the expectation is that 
the customer can take the new gear, put in his old credentials and come 
up on the new services.

The DSL networks are very, very large and very slow to migrate and the 
answer to Iljitsch's comment, they do not move unless they take the new 
service and the new gear.  The whole point of the design is to allow a 
migration from PPP to DHCP Auth without any touch to the AAA databases 
or requiring a big flag day for million user networks.

Regards,
Ric

Ted Lemon wrote, around 1/11/07 9:58 AM:
> On Oct 29, 2007, at 12:16 AM, Stig Venaas wrote:
>
> [...]
>
> At Ralph's request, I've gone back and read the archives for this 
> discussion.   Much learned and useful argument has been presented, 
> including Stig's most recent objection, which is certainly a good 
> valid technical objection.
>
> However, what I have *not* noticed, which I find kind of stunning, is 
> anybody from the "let's do it" side of the argument answering 
> Iljitsch's very cogent and telling objection here:
>
> http://www1.ietf.org/mail-archive/web/int-area/current/msg01014.html
>
> This objection is a showstopper.   If it can't be answered, there is 
> no chance that this protocol can ever be deployed.   All the technical 
> and procedural objections that have been discussed here are irrelevant 
> if this one objection can't be answered.
>
> So I think that before any further discussion on the merits of the 
> wire protocol go forward, perhaps it would behoove this learned body 
> to actually address Iljitsch's objection.   I must confess that I have 
> not read every message in the archive, because there are a remarkable 
> lot of them, so if someone can point me at the place where Iljitsch's 
> objection was shown to be unnecessary, I would consider it a great favor.
>
>
>
> _______________________________________________
> dhcwg mailing list
> dhcwg at ietf.org
> https://www1.ietf.org/mailman/listinfo/dhcwg

_______________________________________________
dhcwg mailing list
dhcwg at ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg