The MPLS WG Archive

["Angela Chiu" <alchiu@research.att.com>]

I meant "thesis topic" in my last email.

Yakov,

I agree with your other point that the market will eventually determine the
viability of the two or more models. From a carrier's point of view, since
the overlay model is most likely the one to be used first, we want to make
sure that key capabilities are supported by the overlay model. In
particular, routing a new lightpath that is diverse from one or more
existing lightpaths is not only needed by the capability that Kireeti
described earlier, i.e., layer-3 like protection switching, but also needed
in cases where an upper layer network wants to have its own survivability.
For example, an IP network wants to have IP layer only restoration (e.g., if
link C-D needs to support additional traffic when link A-B fails, they need
to be diversely routed), or a private line network needs to have diversity
among a set of links as John pointed our earlier (for more details on
diverse routing, see the draft we presented in last IETF,
http://search.ietf.org/internet-drafts/draft-chiu-strand-unique-olcp-00.txt,
a new version will be issued in a few weeks).

As Kireeti pointed out before, certain extensions to the existing GMPLS
protocols needed to provide this type of diversity capability is currently
not defined yet, i.e., messages that a client uses to request a new
lightpath that is diverse from one or more existing lightpaths. Since this
requirement has been defined in OIF carriers' requirement document as well
as our draft, I think we can all work together to make it happen.

Regards,

Angela



-----Original Message-----
From: owner-mpls@UU.NET [mailto:owner-mpls@UU.NET]On Behalf Of Angela
Chiu
Sent: Wednesday, October 25, 2000 10:03 AM
To: 'Yakov Rekhter'
Cc: ip-optical@lists.bell-labs.com; mpls@UU.NET; sc@tellium.com;
xuyg@lucent.com; yxue@UU.NET
Subject: RE: [IP-Optical] RE: Optical link bundling. Was Re:
DraftMinutes From Pittsburgh


Yakov,

Yes, you are right. Routing the primary and backup paths jointly is always
more optimal than fixing the path for primary first then routing the backup
accordingly. The same argument can be applied to comparing centralized
routing with distributed routing. Even distributed routing is less optimal,
it is the trend today. I think the real issue is at what cost the additional
optimality is gained, and how much the additional optimality is in a typical
network setting. In this case the cost is all the topological information
including SRLG information as well as physical impairment constraints in the
optical network that routers need to obtain in order to make proper routing
decision.

I have an idea, this will be a valid master/PhD thesis for some graduate
students who would like to work on real world problems.

Regards,

Angela

-----Original Message-----
From: ip-optical-admin@lists.bell-labs.com
[mailto:ip-optical-admin@lists.bell-labs.com]On Behalf Of Yakov Rekhter
Sent: Wednesday, October 25, 2000 7:35 AM
To: alchiu@research.att.com
Cc: ip-optical@lists.bell-labs.com; mpls@UU.NET; sc@tellium.com;
xuyg@lucent.com; yxue@UU.NET
Subject: Re: [IP-Optical] RE: Optical link bundling. Was Re:
DraftMinutes From Pittsburgh


Angela,

> Some followup discussions in line.

more in line...

> Regards,
> Angela
>
> -----Original Message-----
> From: owner-mpls@UU.NET [mailto:owner-mpls@UU.NET]On Behalf Of Kireeti
> Kompella
> Sent: Monday, October 23, 2000 1:58 PM
> To: kireeti@juniper.net; sc@tellium.com; xuyg@lucent.com; yxue@UU.NET
> Cc: ip-optical@lists.bell-labs.com; mpls@UU.NET
> Subject: RE: [IP-Optical] RE: Optical link bundling. Was Re:
> DraftMinutes From Pittsburgh
>
>
> > I don't see why TE and protection require the routers to specify
explicit
> > routes.
> > The routers can simply specify to the optical layer what type of optical
> > layer protection
> > it requires.
>
> Suppose router A wants to get to router B, and wants to take two
> different ingress and egress points in the optical domain, X->Y
> for the primary LSP, and W->Z for the backup.  A does not require
> optical protection for the X->Y path, nor for the W->Z path.  A
> *does* require that the X->Y path and the W->Z path do not share
> common links.  How is this to be done?
>
> If A did the full path computation, this is simplicity itself.
>
> [AC] I think you have a good point here. I also heard the same kind if
> reasoning (i.e., have a layer-3 like protection switching) for supporting
> the peer model. But after discussing with others, it seems that overlay
> model should be able to provide the same capability.

Not really... for more on this see below...

> Normally, the primary
> LSP X->Y is set up first, and becomes a forwarding adjacency (FA)
according
> to your LSP Hierarchy draft. Then the associated information of the FA
X->Y
> including its exact path and SRLG information should be propagated via IGP
> extensions, same as with any other link in the network. Thus if router A
> sends a request to OXC W to set up a backup lightpath from W->Z to be
> diversely router from the existing FA X->Y, OXC W should already have the
> right information to perform proper routing.

It is a known fact that for computing disjoint paths the approach
you outlined above may result in a situation where no backup
path will be found, despite the fact that that it is possible
(using some other approach) to find two disjoint paths.

> Comparing with the peer model solution where routers need to know all the
> SRLG information of the optical domain as well as all relevant physical
> impairments in the optical signal in the case of transparent optical
> network, it is still not clear to me which one is simpler.
>
> I think it is very good to have this kind of technical discussion openly
on
> the list. Hope others can provide more technical and business (after all
> carriers need to pay for these features) evidences for the need of each
> model. Some other reasoning I heard includes that peer model can improve
the
> IGP scalability in terms of the number of neighbors a router needs to peer
> with. But since large ISPs today seem to cope well with the IGP
scalability
> today, I don't see why the problem will get significant worst when optical
> networks come into play.

In the end it is not the discussion on this list, but the competition
in the marketplace that will determine the viability of different
models.

Yakov.

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