The MPLS WG Archive

["DECRAENE Bruno RD-CORE-ISS" <bruno.decraene@francetelecom.com>]

David,

Thanks for the feedback.

I'll need to re-read LSP Ping to check if there is a potential
interaction. Could you please be more specific on the LSP Ping section
that could be impacted?

However, please note that the LDP FEC is not changed on its way from
egress to ingress. In particular, the LDP FEC is not aggregated and all
LSRs on the way know the exact same FEC.

Cheers,
Bruno

> -----Original Message-----
> From: David Allan [mailto:dallan@nortel.com]
> Sent: Thursday, April 13, 2006 4:29 PM
> To: ewgray@GraIyMage.com; DECRAENE Bruno RD-CORE-ISS;
pdutta@riverstonenet.com
> Cc: mpls@ietf.org
> Subject: RE: [mpls] One LDP Implementation specific question ofreceive
labelmapping for prefix FECs
> 
> FYI
> 
> It does go further, re-writing FEC stack processing for LSP ping to
permit interpretation and
> matching of more specific to less specific prefixes and figuring out
whether that was actually
> permissable in the local context (a.k.a. more specific prefix did not
exist at the terminating LSR)
> could get truly ugly...
> 
> cheers
> Dave
> 
> > -----Original Message-----
> > From: Eric W Gray [mailto:ewgray2k@netscape.net]
> > Sent: Thursday, April 13, 2006 10:17 AM
> > To: DECRAENE Bruno RD-CORE-ISS; pdutta@riverstonenet.com
> > Cc: mpls@ietf.org
> > Subject: Re: [mpls] One LDP Implementation specific question
> > ofreceive labelmapping for prefix FECs
> >
> >
> > Bruno/Pranjal,
> >
> >     One thing I would like to make perfectly clear in this
> > discussion is
> > that - if either or
> > both of you would like to write an ID that proposes changing or
> > extending LDP to do
> > what you suggest, you should do so.  The suggestions you make may be
> > well received
> > by participants in the MPLS working group.
> >
> >     However, you should not view this as something to
> > "correct" in the
> > current LDP
> > specification.  The point of the current LDP specification work is
to
> > document what is
> > being done in current interoperable implementations, now.
> >
> >     I believe the point has been sufficiently well made by
> > now that the
> > text - as is -
> > represents current implementation better than the changes
> > you've implied
> > or explicitly
> > suggested should be made with respect to the exact match
> > requirement in
> > handling an
> > arriving label distribution.
> >
> > --
> > Eric
> >
> >
> > _______________________________________________
> > mpls mailing list
> > mpls@lists.ietf.org
> > https://www1.ietf.org/mailman/listinfo/mpls
> >
> >

_______________________________________________
mpls mailing list
mpls@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/mpls