The MPLS WG Archive[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] [IP-Optical] RE: Optical link bundling. Was Re: Draft Minutes From Pittsburgh
Seems to me, Darren, that most in the IETF _do_ expect the OTN to have primarily one client: IP. This would also track with other efforts, such as diffserv, enum, and VoIP, in which any non-IP-centric signaling is relegated to the edges (or up in the layer structure, such as H.323). Wasn't the biggest problem with ATM deployment that its clients were almost exclusively IP, while its signaling structure was absolutely not? The early adopters were not what had been intended by design, IMO. In the late 1980s, maybe traffic loads were not favoring IP. The common wisdom now is that they do favor IP. Bert albert.e.manfredi@boeing.com > -----Original Message----- > From: darren.freeland@bt.com [mailto:darren.freeland@bt.com] > > Hi John, > > At the risk of sounding repetitous ... > > >> There's no reason why the back > >> end couldn't be GMPLS in peer mode > > I believe there is a reason. And I've yet to be convinced > otherwise by > anyone. The way I see it, peer model = single control plane. Single > control plane = common addressing, signalling, & routing. > Common addressing > = only possible if the OTN has a single client. This has > been explained in > the recent "server layer trails = client layer links" links > discussions => > i.e. common addressing not possible in a multi-client OTN => > i.e. peer model > impractical in a multi-client OTN. I think this fundamental > issue has to be > recognised. > > Regards, > Darren. |
|