RE: WG Consensus Call: draft-swallow-gmpls-overlay-00.txt

[John Drake <jdrake@calient.net>]

Comments in line

-----Original Message-----
From: Bala Rajagopalan [mailto:BRaja@tellium.com]
Sent: Wednesday, December 11, 2002 12:03 PM
To: John Drake; 'Lazer, Monica A, ALASO'
Cc: ccamp@ops.ietf.org
Subject: RE: WG Consensus Call: draft-swallow-gmpls-overlay-00.txt


JD:  A general comment is that I think your e-mail is sort of orthogonal
to my e-mail.  I.e., I'm not sure how what you wrote relates to what I
wrote.


John,

I don't agree with your assertions below.
The swallow draft addresses the case where 

1. it's implicit that you're running GMPLS RSVP-TE within the core


JD:  This is assumed, but not required.


2. you're using IP addresses (only) for transport NEs


JD:  Are you saying that you cannot signal IP addresses in UNI 1.0?  This
is surprising.


3. there's a possibility of exchanging some (limited)
   reachability information between the core and edge


JD:  Reachability exchange is mentioned in the draft.  It is certainly
is not required, in which case the functionality reduces to UNI 1.0.


As I mentioned earlier, I think it's useful to 
consider this specific scenario further.

But these assumptions are not made in the case of UNI 1.0.


JD:  Fine, but what does any of this have to do with my e-mail?


Bala


> -----Original Message-----
> From: John Drake [mailto:jdrake@calient.net]
> Sent: Wednesday, December 11, 2002 12:31 PM
> To: 'Lazer, Monica A, ALASO'; Kireeti Kompella; Khuzema Pithewan
> Cc: ccamp@ops.ietf.org
> Subject: RE: WG Consensus Call: draft-swallow-gmpls-overlay-00.txt
> 
> 
> JD:  This is simply not true.  Since we're dealing with an 
> overlay model,
> any control plane can be used inside the cloud.
> 
> 
> UNI 1.0 allows transition to use of GMPLS-based protocols
> without the assumption of overnight replacement of existing control
> plane implementations, which is not even feasible for a large network.
> 
> 
> JD:  You seem to be implicitly asserting that this is NOT the 
> case with
> this interface.  If this is in fact what you are asserting, then it is
> simply not true.
> 
> 
> Monica A. Lazer
> Network Architecture and Reliability
> 
> 908 234 8462
> mlazer@att.com
> 
> -----Original Message-----
> From: Kireeti Kompella [mailto:kireeti@juniper.net]
> Sent: Wednesday, December 11, 2002 5:04 AM
> To: Khuzema Pithewan
> Cc: ccamp@ops.ietf.org
> Subject: RE: WG Consensus Call: draft-swallow-gmpls-overlay-00.txt
> 
> > This draft addresses the overlay networks which UNI1.0 from OIF
> addresses.
> > shouldn't we have one document addressing one problem. In 
> fact, UNI1.0
> > addresses it in-depth, so why there is a need of another document
> talking
> > about overlay networks?
> 
> I guess you haven't really been following the CCAMP email, so I'll
> repeat:
> (a) the OIF UNI1.0 is *not* a standard;
> (b) just because the OIF is discussing something doesn't preclude the
>     IETF from discussing it;
> (c) the problems addressed are not quite the same.
> 
> If having two documents in this space is an issue, please make your
> concerns known to the OIF.
> 
> Kireeti.
>