[nsp] MPLS/LSP question

[Eric Osborne]

On Wed, Feb 19, 2003 at 03:58:19PM +0200, Hank Nussbacher wrote:
> Consider the following topology:
> R1-R2-R3
> Where R1,R2 are MPLS routers. R3 doesn't run MPLS.
> 
> Can I built an LSP that ended in R3 without supporting MPLS? According to 
> RFC3031 it seem that it is possible:
> "The definition of LSP allows for the LSP Egress to be a node which
> does not support MPLS; in this case the penultimate node in the LSP
> is the Proxy Egress."
> 
> However, I am not sure, if it indeed supported defacto, and if yes, please 
> send sample IOS config.

Semantics.  The LSP doesn't "end on" R3 insofar as R3 is not
explicitly involved in any MPLS signalling.  However, you can
certainly have packets go from R1 to R2 to R2 which are labeled
between R1 and R2 and not labeled when they get to R3.  The term
"Proxy Egress" isn't one that really gets bandied about in casual
conversation; just turn on LDP between R1 and R2 and it'll probably do
what you want.  But what specifically do you want to do?

Config depends on code version.  On newer code, you want to use LDP
and the 'mpls' syntax.  Globally, configure 'mpls label protocol ldp'
and then 'mpls ip' on the interface between R1 and R2 - this goes both
on R1 and R2.

On older code, turn on 'tag-switching ip' on the interface between R1
and R2, and TDP will be turned on.  Functionally equivalent protocols,
but use LDP if you can.

Again, this will probably do what you're asking for, but I'm not
convinced it's all that useful in and of itself - what's the end goal
here?



eric

> 
> Thanks,
> Hank
> 
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> http://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/