The MPLS WG Archive

[Alia Atlas <aatlas@avici.com>]

The draft doesn't discuss at all the case where packets may be distributed 
to different paths based upon the MPLS label stack.   This can occur for 
some equipment when either the packet is not identified as IPv4 or IPv6 or 
when the equipment cannot examine the header underneath the MPLS label 
stack, presumably due to its depth.

This doesn't cause concerns for the PWE3 cases, but it was a definite issue 
when we were discussing LSP ping and other OAM issues.  For instance, this 
was part of the concern for adding an OAM label into the label stack.

If this is really to be a BCP describing how ECMP and MPLS interact, then 
I'd really like to see a very short section describing the possibility that 
packets may be load-balanced based on the label stack as well.

At the time of the OAM label discussion, it was suggested that future 
implementations shouldn't include the reserved labels in any processing of 
the label stack done for load balancing.   Perhaps we don't want to capture 
that here - the idea seemed a bit sensitive at the time - though I'd be in 
favor of it.

Alia

At 01:16 PM 2/13/2005, George Swallow wrote:
>This begins a two week last call on
>
>    Avoiding Equal Cost Multipath Treatment in MPLS Networks
>
>       <draft-ietf-mpls-ecmp-bcp-00.txt>
>
>    The last call ends Feb 27 24:00 UTC.
>
>...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



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