[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: ASON reqts

lyndon,

also some comments in-line...

Ong, Lyndon wrote:
Hi Folks,

I'm hoping that the requirements draft does not imply that we will now define an alternative to 3474, that was not my
intention when participating in the draft. I see no reason to rethink what was done in 3474, 
well i think there where listed in the signalling proposal
(take a close look at section 3.1):
http://www.ietf.org/internet-drafts/draft-dimitri-ccamp-gmpls-rsvp-te-ason-00.txt

i am not sure that trying to "impose" a solution that breaks
the gmpls common developments is a *good* solution
which has now been implemented and tested by a number of participants
It would be good to understand if people in the group are
interested in supporting ASON, 
i think one of the good things that this document delivers
is that it clearly states what "ason support means" from the
signalling perspective, the abstract of this i-d seems (since
so far) to be very clear for the informed reader:

"This document concentrates on the signaling aspects of the GMPLS
 suite of protocols. It identifies the features to be covered by the
 signalling protocol to support the capabilities of an Automatically
 Switched Optical Network (ASON). This document provides a problem
 statement and additional requirements on the GMPLS signaling
 protocol to support the ASON functionality."


however, which seemed to be
a controversial issue on the mailing list a while back.
Also, as Stephen points out, it would be good to get
feedback from ITU SG 15 on the draft, as it does represent
only the authors' interpretation of ASON requirements and
may not be complete.
see the response from an itu participant:
"The document is very aligned with the ITU work. From an ITU perspective, the document is easy to read. And considering the author list, it's a balanced list of ITU and IETF participants. Following up on Adrian's request, I would be interested to hear from IETF participants. As Kireeti stated, this document is an attempt to bridge the gaps between the IETF and ITU work. Most important for progressing this work is the understanding level/interest by the IETF community involved in GMPLS."

but since you're looking for more formal ITU review, you know
very well that such things happen much more easily when the i-d
as a WG status.

thanks,
- dimitri.

Cheers,

Lyndon

-----Original Message-----
From: Stephen Trowbridge [mailto:sjtrowbridge@lucent.com]
Sent: Monday, May 12, 2003 3:23 PM
To: John Drake
Cc: 'Wijnen, Bert (Bert)'; Kireeti Kompella; ccamp@ops.ietf.org
Subject: Re: ASON reqts


John,
Ahhhh....
This sounds like a whole new can of worms. The goal therefore does
seem to be to develop yet another protocol solution, and I think
we need to be VERY careful before taking this kind of step.

Your assertion now seems to be that RFC 3473+3474 does NOT meet the
ASON requirements. If this is the case:
- Does G.7713.2 meet the requirements, and we missed something in
  the translation to RFC 3474? If so, then we need to supercede
  RFC 3474 with a better translation, not start over.

- Do you think that even G.7713.2 does not meet the requirements?
  If you believe this to be the case, it seems like the obvious
  first step would be to inform ITU-T - after all, this is where
  the ASON requirements came from and it makes no sense to start
  a new protocol without fixing (and aligning) G.7713.2.

If we made a mistake, lets fix it, but lets NOT start proliferating
ASON extensions take 2; ASON extensions take 3; ...

Regards,
Steve

John Drake wrote:


I think that the problem is that RFC 3474, in addition to its technical
issues, doesn't address all of the ASON requirements.  For example, it does
not address full call/connection separation.  So we're following the post
3474 procees of documenting the ASON requirements relevant to GMPLS
signalling in order to have a common understanding of what problem is to be
solved.



-----Original Message-----
From: Stephen Trowbridge [mailto:sjtrowbridge@lucent.com]
Sent: Monday, May 12, 2003 2:27 PM
To: John Drake
Cc: 'Wijnen, Bert (Bert)'; Kireeti Kompella; ccamp@ops.ietf.org
Subject: Re: ASON reqts


Bert & John,
I think this document has nothing to do with the issue Bert
mentioned. This was a protocol issue, not a requirements issue.
A decision was made to leave the message out of 3474 with the
belief that:
- The protocol works without it; and
- The presence of the message violates the protocol.
The appropriate next step here would be a liaison back to ITU-T
explaining what was done and why, and if the ITU-T agrees with
the reasoning, they can align their protocol document (in this
case, G.7713.2).

Back to Bert's original questions:


- is there or do we see any conflict?



- As Zhi said, this is not (yet) fully aligned with ITU-T requirements
 and should not be advanced to a WG document until it is.
However, before
 we do, I think we need to understand why this is necessary
- see below.





- are we duplicating some work?





- Almost certainly. If we take the kind of alignment with ITU-T
 requirements that Zhi mentions as a "gate" for progressing
this, then
 what is in this document that is different from G.8080 and
G.7713 and
 why do we need this in place of a normative reference?





- what is the purpose of this draft?





Excellent question.




 - is it after the fact documenting of requirements?





If it is, why bother? Also, we already have the requirements
(developed
before the protocol) in G.8080 and G.7713.





 - is it getting ITU-T documented requirements in RFC form?





This could be a reasonable goal, if this document is to be an info
track requirements document characterizing G.8080 and G.7713 for
IETF folks in the same way that RFC 3474 does for G.7713.2 and
RFC 3475 does for G.7713.3. But it is hard to understand why producing
an IETF translation of the ITU-T question would be very high priority
when we already have an IETF translation of the answer, but
fundamentally
there would be no problem if this were the objective.





 - is it extending ITU-T documented requirements?





I don't think so. If it is, I hope we start with a liaison rather
than trying to extend ITU-T requirements without talking to
ITU-T first.





 - is it contrdicting them?





This is my biggest worry. I don't think we should start the process
over again and open the door that we come up with yet another
protocol solution to address the same requirements. Vendors are
already building to G.7713.2, G.7713.3 (and by extension, to
RFC 3473+3474 and RFC 3472+3475). We most definitely should NOT
start a new work item with an objective to develop a new protocol
solution to the same problem.





 - is it meant to be used as communication to ITU?





I haven't seen this proposed, but I think that if the goal is
to capture ITU-T requirements in RFC form, it would be a good
idea to ask ITU-T whether the proposed document accurately
captures their requirements.

/Steve


John Drake wrote:




Bert,

I think so.

Thanks,

John





-----Original Message-----
From: Wijnen, Bert (Bert) [mailto:bwijnen@lucent.com]
Sent: Monday, May 12, 2003 1:27 PM
To: John Drake; Kireeti Kompella; ccamp@ops.ietf.org
Subject: RE: ASON reqts


I know that some people had issues with the way that ITU-T
had defined the RSVP-TE extensions for ASON. In fact there
was/is a claim that it is broken. So we removed the offending
text from the RFC3474.

The next thing we were going to do (as far as I understood it)
is to document why we (or some of us in IETF) think that the
ITU-T solution is broken, potentially with suggested fixes.
That we would send to ITU-T SG15.

Is this the first step of that process?

Thanks,
Bert





-----Original Message-----
From: John Drake [mailto:jdrake@calient.net]
Sent: maandag 12 mei 2003 22:24
To: 'Wijnen, Bert (Bert)'; Kireeti Kompella; ccamp@ops.ietf.org
Subject: RE: ASON reqts


Bert,

I thought that the post 3474/3475 process started with a



requirements




document.

Thanks,

John





-----Original Message-----
From: Wijnen, Bert (Bert) [mailto:bwijnen@lucent.com]
Sent: Monday, May 12, 2003 1:14 PM
To: Kireeti Kompella; ccamp@ops.ietf.org
Subject: RE: ASON reqts


So if we look at RFCs 3474/3475 and the ITU-T documents
that those 2 RFCs point to, then I wonder:
- is there or do we see any conflict?
- are we duplicating some work?
- what is the purpose of this draft?
 - is it after the fact documenting of requirements?
 - is it getting ITU-T documented requirements in RFC form?
 - is it extending ITU-T documented requirements?
 - is it contrdicting them?
 - is it meant to be used as communication to ITU?

Just wondering what is happening here.

Thanks,
Bert





-----Original Message-----
From: Kireeti Kompella [mailto:kireeti@juniper.net]
Sent: maandag 12 mei 2003 17:25
To: ccamp@ops.ietf.org
Subject: Re: ASON reqts


Hi All,

On Fri, 2 May 2003, Kireeti Kompella wrote:





To take things one at a time, it would be very useful to



read and comment




on the ASON reqts draft, as this was deemed tremendously



important, and




a rich source of misunderstanding and cross-talk; and



coming to a common




understanding over this should help get the IETF and the





ITU working




together.



I haven't seen many comments, so the assumption is





either that no




one cares, or that folks have read it and have no issues.

I'd like to get a reading on whether this doc is





ready to be a




CCAMP WG document.  Please respond (preferably publicly)



with one of:




- "I have read this document and it is ready to be a CCAMP



WG doc" OR




- "I have read this document, and it isn't ready to be a



CCAMP doc".




Note that if there aren't enough responses, the default



assumption is




that the document is either not of interest or not





ready, and in




either case will not become a CCAMP WG doc.  Note too





that this doc




is an attempt to bridge some gaps between the IETF and





the ITU-T,




and as such is fairly important.  It would be





useful to give an




update on its status at the interim T1X1 meeting in June.

Please get your responses in by COB Friday May 16th.

Thanks,
Kireeti.











--
Papadimitriou Dimitri
E-mail : dimitri.papadimitriou@alcatel.be
Private: http://www.rc.bel.alcatel.be/~papadimd/index.html
E-mail : dpapadimitriou@psg.com
Public : http://psg.com/~dpapadimitriou/
Address: Fr. Wellesplein 1, B-2018 Antwerpen, Belgium
Phone  : +32 3 240-8491