The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2005-Mar> msg00015



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

[mpls] Last Call on BCP on ECMP

  • From: Alia Atlas <aatlas@avici.com>
  • Date: Thu, 03 Mar 2005 13:41:39 -0500
  • Cc: mpls@ietf.org
  • X-Avici-MailScanner-Information: Please contact the ISP for more information

Hi Neil,

The BCP is trying to describe what (generally) currently happens and the 
ramifications for ECMP.

Any reserved label, no matter where it is located in the stack, may be 
included in the computation for load balancing based on the stack.  This 
causes traffic using such reserved labels to not necessarily fate share 
with traffic that doesn't.  Examples are traffic with a router alert label 
or with an OAM label.

In this BCP, I think we just want to capture the implications of this 
behavior and not, explicitly, say how to do it better - because most 
current practices do not.

Alia

At 11:53 AM 3/3/2005, Allan, David [CAR:NS00:EXCH] wrote:

>Hi Neil:
>
>IMO if a reserved label defines a function and not forwarding, then it is 
>likely that we can envision future functions whereby they should fate 
>share with the LSP of interest, a far more useful behavior than not fate 
>sharing. Frequently they would be either bottom label or nested beneath 
>the bottom label...but I would not cast that in stone (witness explicit 
>null ;-)
>
>Fate sharing is not the case today....
>
>Dave
>
> > -----Original Message-----
> > From: neil.2.harrison@bt.com 
> [<mailto:neil.2.harrison@bt.com>mailto:neil.2.harrison@bt.com]
> > Sent: Thursday, March 03, 2005 10:39 AM
> > To: aatlas@avici.com; swallow@cisco.com
> > Cc: mpls@ietf.org; dallan@nortelnetworks.com
> > Subject: RE: [mpls] Last Call on BCP on ECMP
> >
> >
> > According to rfc3032 labels 0-15 are reserved...so its seems
> > there is a more general point here.  Further, there seem to
> > be proposals for some of these to occur at arbitrary
> > positions in the stack, eg draft-ietf-mpls-explicit-null-02.txt.
> >
> > So how does this play here?
> >
> > regards, Neil
> >
> > > We may want another or a different example of a reserved
> > > label (such as the
> > > Router Alert or even Explicit Null)
> > >
> > > Alia
> > >
> > > At 08:07 PM 3/2/2005, George Swallow wrote:
> > > > > "The processing of the label stack for load balancing
> > may include
> > > > > all MPLS labels, including those that are reserved; in
> > > such a case,
> > > > > the addition of a reserved label, such as the OAM label,
> > > may result
> > > > > in a different path being taken."
> > > >
> > > >I'll probably word-smith it a bit, but I'm happy to capture the
> > > >thought.
> > > >
> > > >...George
> > > >
> > > >=============================================================
> > > ===========
> > > >George Swallow             Cisco Systems
> > > (978) 936-1398
> > > >                            1414 Massachusetts Avenue
> > > >                            Boxborough, MA 01719
> > >
> > >
> > >
> > > _______________________________________________
> > > mpls mailing list
> > > mpls@lists.ietf.org 
> <https://www1.ietf.org/mailman/listinfo/mpls>https://www1.ietf.org/mailman/listinfo/mpls 
>
> > >
> >
> >



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