Re: [dhcwg] Call for adoption: draft-korhonen-dhc-pd-exclude

Hi,

I am the initial author of draft-savolainen-stateless-pd.. the biggest issues with that approach that I recall were:
1) It is another solution for the same problem
2) It would likely take too long time to make it to 3GPP Rel-10 (or even 11..)
3) The provisioning might require use of DHCPv6 client/server anyway (for OPTION_6SPD)
4) With stateless PD there is bigger risk that client and server somehow get out of sync (with stateful approach a bug in DR does not necessarily damage much, but in stateless approach it would more easily result in out of sync situation)
5) the DHCPv6 PD can be implemented to provide very similar solution:
* having lightweight implementations of DR at the first hop gateway and RR at the node
* gateway selecting prefixes for delegation exactly at the manner the Stateless PD described (for more efficient routing)
6) The people are already fixed on DHCPv6 PD and are not very eager to listen other options:)
7) Eight years too late?-) (Much of discussion leading to DHCPv6 PD seemingly took place at around 2002)

I think there was something else as well, which I don't recall immediately..

This DHCPv6 PD Exclude prefix option actually helps a lot to get benefits we sought with "stateless PD" out of stateful DHCPv6 PD domain.

Best regards,

Teemu

P.S. I also didn't manage to find much (but found some) support to the stateless approach in previous IETFs, but thank you for the support now:) 

> ----- Original Message ----
> > From: Ted Lemon <Ted.Lemon@nominum.com>
> > To: Behcet Sarikaya <sarikaya@ieee.org>; Behcet Sarikaya
> ><behcetsarikaya@yahoo.com>
> > Cc: Ole Troan <otroan@employees.org>; "dhcwg@ietf.orgGroup"
> <dhcwg@ietf.org>
> > Sent: Mon, September 27, 2010 4:54:05 PM
> > Subject: Re: [dhcwg] Call for adoption: draft-korhonen-dhc-pd-exclude
> >
> > On Sep 27, 2010, at 1:24 PM, Behcet Sarikaya wrote:
> > > There is simply no  need for such a thing :-).
> >
> > This is kind of a weird objection.    Clearly the authors of the
> draft have
> >some reason for building the thing the  draft documents.   Perhaps it
> would be
> >more fruitful to discuss whether  this is the right way to address the
> use case
> >they are addressing with this  draft, rather than simply dismissing
> whatever
> >their use case is as  nonexistent?   I mean, if they don't actually
> *have* a use
> >case, then your  point would prevail, but I assume they do or they
> wouldn't have
> >bothered to make  the proposal!
> >
> 
> OK, using your logic, the same set of authors has made a draft about
> stateless
> PD. They did bother making one. It is at:
> http://tools.ietf.org/id/draft-savolainen-stateless-pd-01.txt in case
> you have
> not seen it.
> 
> Now they are making this proposal which is kind of the reverse
> proposal.
> 
> Can you please point out why you are picking up this one and not the
> other one?
> 
> The point is that authors make all sorts of proposals. It is up to the
> WG to
> advance it or not.
> 
> 
> 
> _______________________________________________
> dhcwg mailing list
> dhcwg@ietf.org
> https://www.ietf.org/mailman/listinfo/dhcwg