The MPLS WG Archive

["AMK Networking Technologies Inc." <amknt@sympatico.ca>]

Hi Chris,

I don't see the requirement for this functionality as LDP offers the
following:

1) Two distribution methods ( DoD and DU)
2) Two control methods (ordered and Independent)
3) Two retention methods (liberal and conservative)

These can be combined to achieve varying results *some not so desirable*,
like break-and-make and make-and-no-break.
If the scenario you describe is like this....a label mapping is received by
a LSR for a FEC which has bound a label, I am assuming that this mapping is
received by that LSR on a different interface. That would also mean that the
network is configured for DU at the very least. The LSR would not
necessarily inform the upstream LSR but rather use the new label and release
the existing label...in that case it is not a break-make situation.

If the scenario you describe is like this....The MPLS domain is configured
for Liberal Label Retention and a new next hop is discovered by a particular
LSR for a particular FEC (the next hop is a different interface...logical or
physical) then the existing label is released and the label stored in the
LIB for that next hop interface is used...again this is not a Break-Make.

If a connection or downstream router fails then that is again a different
set of procedures depending on the configuration of the LSRs in the MPLS
domain. In that case DoD with conservative retention mode does cause a
break-make...life is full of trade offs.

Please feel free to comment...If I misunderstood you, I apologize, please
restate your question.

regards,

Micheal
-----Original Message-----
From: owner-mpls@UU.NET [mailto:owner-mpls@UU.NET]On Behalf Of Chris
Kuszmaul
Sent: Tuesday, May 29, 2001 6:45 PM
To: David Charlap; mpls@UU.NET
Subject: LDP requires break-before-make



 Before a label can be mapped to an FEC that has already been mapped, the
LSR
must release the old label (prompted perhaps by a withdrawal message). Thus,
LDP
calls for a break-before-make on LSP updates.  It would seem a
straightforward
matter for the protocol to include a withdraw-and-map message that performs
this
sequence atomically, and so preserves a make-before-break condition.

  Is there some reason for not providing for this kind of functionality?

 Thanks for any comments.

CLK

PS: The careful reader will realize that I have mentioned this issue before
in
the thread [Re: Receive Label Receive !!], but have not seen any reply. This
is
an attempt to focus the question and broaden the pool of possible
respondents.