The MPLS WG Archive

[Yoshihiro Ohba <yohba@tari.toshiba.com>]

Jim,

From: "James R. Leu" <jleu@mindspring.com>
Subject: Re: Address Message in LDP
Date: Wed, 7 Jun 2000 13:14:29 -0500

> On Wed, Jun 07, 2000 at 11:33:26AM -0400, Bob Thomas wrote:
> 
> <snip>
> 
> > I think you are making this more complicated that it need be.
> >
> > The purpose of the Address Message is to make it possible for an LSR
> > to establish a mapping between the first 4 octets of a peer's LDP Id
> > and addresses bound to the peer.
> > 
> > This mapping is useful in determining whether the peer is the next
> > hop for a particular prefix.
> 
> I understand your second statement (the first about me making it more
> complicated then it need be is always true :-).  With respect to the third
> statement, are you saying that this mapping is or is not the ONLY way to
> figure out if a peer is the next hop for a particular prefix?
> 
> My goal is to get an answer about "non-Directly Connected LSRs" and
> how they guarantee that they are distributing labels and installing them
> on the correct interfaces.
> 
> The way to get the above answer may involve or mimic the way we deal with
> mappings over parallel links between two directly connected LDP peers.
> 
> If I have LDP sessions over parallel links and a particular prefix is
> preferred over only one link, the peer closer to egress may provide mappings for
> this prefix via both sessions.  If I use the address list as my definitive
> source for which session is the next hop, both sessions will match.
> 
> What other information do we use in this case to decide whether or not
> a particular session is ACTUALLY the next hop for a particular prefix?
> 

In the case of parallel links between directly connected LDP peers,
there is another method to obtain the correct mapping between the next
hop and a session:

1. Find an interface that matches the next hop.

2. Find an LDP peer that matches the interface found in step 1.
   Since the interface is on a point-to-point link, at most one LDP
   peer will match that interface.  Note that the mapping between
   an LDP peer and an interface is determined during LDP Link Hello
   message exchange.

I think this method is useful in the case where there is a possibility
that two or more LDP peers on the same neighboring LSR may advertise
overlapping address lists in their Address messages.  Of course, this
is not necessaly if the implementation is good (or complicated?) 
enough to advertise non-overlapping address lists.

Yoshihiro Ohba

> Jim
> -- 
> James R. Leu