The MPLS WG Archive

["Lazer, Monica A, NNAD" <mlazer@att.com>]

See below

Monica A. Lazer
Advanced Transport Technology and Architecture Planning

908 234 8462
mlazer@att.com


 -----Original Message-----
From: 	Yakov Rekhter [mailto:yakov@cisco.com] 
Sent:	Tuesday, October 24, 2000 1:58 PM
To:	neil.2.harrison@bt.com
Cc:	David.A.Holmes@disney.com; Mark.Jones@mail.sprint.com;
ip-optical@lists.bell-labs.com; mpls@UU.NET; kireeti@juniper.net;
sc@tellium.com; xuyg@lucent.com; yxue@UU.NET; zwlin@lucent.com
Subject:	Re: [IP-Optical] RE: Optical link bundling. Was Re:
DraftMinutes From Pittsburgh 

Neil,

[clipped...]

> 	(i)	are traditional IP control-plane facets (so that is, for
> example, v4 addressing, RSVP signalling and a IGP) the correct choice for
an
> OTN?....though to be honest no-one it seems dare raise this most basic of
> questions too loudly;


[MAL] Neil is right. I have not seen any overwhelming evidence as to why a
protocol used for routing individual packets is the best possible
alternative to be used to set-up circuits.

If you have "the correct choice" for an OTN, which is other than 
GMPLS, please share it with the rest of us (by the way, don't
forget to include detailed description of why your "correct choice"
is any better than GMPLS).


> 	(ii)	irrespective (in principle if not practice) of the choice of
> control-plane facets, can these be unified/shared over all network layers?
> 	I don't think we have 'consensus' to either of these as yet (though
> some have aleady made up their minds on part (i), and assume (ii)
follows).

There is no need to reach consensus on (ii) in the MPLS WG, as with
GMPLS the choice between the peer and the overlay model is
up to each service provider. ...{MAL] As long as the OTN will be able to
support IP, ATM, and other types of clients (including the relevant naming
schemes - IP used by IP clients, E.164 used by ATM switches, etc),  third
party signaling, scheduling, and if the routing decisions are within the
OTN, not made by a client router.


Several service providers have specifically asked that the client model be
supported for OTN, in multiple standards bodies/forums.

Yakov.

_______________________________________________
IP-Optical mailing list
IP-Optical@lists.bell-labs.com
http://lists.bell-labs.com/mailman/listinfo/ip-optical