The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2001-Dec> msg00279



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

MPLSOAM BOF meeting draft minutes

  • From: Shahram Davari <Shahram_Davari@pmc-sierra.com>
  • Date: Tue, 18 Dec 2001 12:06:18 -0800
  • Cc: Ping Pan <pingpan@juniper.net>, "Ash, Gerald R (Jerry), ALCTA" <gash@att.com>, mpls@UU.NET

> I'm just arguing that it is of no consequence.  You haven't
> established that there is any consequence so you accuse me of not
> understanding that there are return paths.

I am saying that you never acknowledged that "TWO" disjoint (they should not be the same) paths
are required for Ping's draft to work. One for ICMP echo response and
the other one for LSP-Ping response.

> 
> So now we understand each other.

I hope so!

>A return paths is required for ICMP
> packets.

Yes.

>The signaling return path is also required to work.

Yes, but there is another requirement that you forgot
to mention: It is required that the MPLS signaling can piggyback
the LSP-ping response at any time, which is only doable for RSVP-TE
because it is based on refreshes. It can't be done with 
other control-planes or static configured LSPs.

> This is
> of no consequence.

People can judge about this. 


> 
> Thanks.  Over and out.

Sure.

-Shahram