The MPLS WG Archive

[Bora Akyol <akyol@pluris.com>]

I would urge you to look at what other network operators/vendors are 
deploying in their networks.

Specifically, what people are doing is running a modified version of the 
IGPs that use an SPF that ALWAYS prefers MPLS LSPs if there is an LSP that 
originates from the router running the SPF to the router that the SPF is 
trying to find a shortest path to. Then this new found route (LSP) is 
inserted directly into the route table.

So you get automatic LSP to FEC binding which does not need to be signaled 
by the Label Distribution protocols (RSVP, CR-LDP, LDP). IMHO, this is nice.

Therefore, I believe that explicitly specifying the FEC and LSP binding is 
only justified in very unique situations; otherwise, it is history.

Henk Smit et al, submitted an ID about this months ago (expired by now). I 
guess it did not quite get read.

Bora Akyol
Pluris (http://www.pluris.com)

At 11:13 AM 4/26/00 -0400, David Wilder wrote:
>Hello,
>
>I don't understand why you say that FEC to LSP bindings would be 'historical'
>in router-based networks.  Would one of you expand on that for me?
>
>
>Thanks,
>
>Dave
>
>
> > At 17:39 04/25/2000 -0700, Bora Akyol wrote:
> > >I presume you are not familiar with how MPLS is actually used in 
> providers networks.
> >
> > I presume you are not familiar with how MPLS is actually used in providers'
> > ATM networks. :)
> >
> > >FEC to LSP binding is almost historical now.
> >
> > That's arguably true in router-based networks, because of the
> > predominance of TE as an application of MPLS in those networks.
> > However, FEC to LSP binding is very much alive in ATM MPLS
> > networks. Also, at least one MPLS-based VPN scheme relies on
> > hop-by-hop routed MPLS (which may, in turn, be over-ridden by
> > TE) in network cores of any type.
> >
> > Regards,
> >
> > Jeremy Larence
> >
> >
> >
> >
> > >At 10:55 AM 4/25/00 +0530, Manish R. Shah. wrote:
> > >
> > >>Hi !!!
> > >>
> > >>Once the LSP's are created, the ingress router looks for the matching 
> FEC, and attaches
> > >>a label corresponding to that FEC on to the data packet. Further 
> forwarding is done using
> > >>the label switching.
> > >>
> > >>In an MPLS domain the LDP has already taken care of mapping
> > >>FEC's with the incoming and outgoing Labels and efficient 
> distribution of Labels.
> > >>
> > >>Look at the following draft for furthur 
> details.  draft-ietf-mpls-ldp-06.txt
> > >>
> > >>Cheers !!!
> > >>
> > >>Manish R. Shah.
> > >>Senior Software Engineer,
> > >>Future Software Pvt Ltd.
> > >>480-481, Anna Salai, Nandanam
> > >>Chennai 600035.
> > >>Phone: +91-(44)-433-0550 Xten 294.
> > >>
> > >>
> > >>-----Original Message-----
> > >>From:   Neetu Gupta [SMTP:neetug@daewoo.dti.daewoo.co.kr]
> > >>Sent:   Tuesday, April 25, 2000 10:28 AM
> > >>To:     mpls@uu.net
> > >>Subject:        Finding the Egress Point in the MPSL Domain for the 
> specific Destination
> > >>
> > >>Hello,
> > >>     Suppose the LSPs are created among the Edge Routers in the MPLS
> > >>Domain with specific class of services. Now the data packet comes at the
> > >>Ingress LER, the IP Header contains the Destination Address and TOS
> > >>byte.
> > >>
> > >>Since there are labels to Egress Router, to which Egress LER the data
> > >>packet will be forwarded that will route to the Destination?
> > >>
> > >>I've heard of the draft that specifies how to treat LSP tunnel end node
> > >>like LSP egress edge as next node for route calculations.
> > >>Please give some reference.
> > >>Neetu
> > >
> > >
> > >
> > >
> >
> >