The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2002-Sep> msg00027



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

2547 BGP VPNs

  • From: Rahul Aggarwal <rahul@redback.com>
  • Date: Tue, 10 Sep 2002 13:14:55 -0700
  • Cc: mpls@UU.NET
  • User-Agent: Mutt/1.4i


On Tue, Sep 10, 2002 at 04:02:32PM -0400, Sachin Kalra wrote:
> Dear MPLS Community:
> 
> Please consider the following simple but strange scenario.

Its not all that strange :) 

> Lets say we have two directly connected PEs. Traffic is flowing from CE1 to 
> CE2. According to VRF table entry on PE1, for destination CE2 bottom label 
> is "X" and top label is "Y". And, we are doing penultimate-hop-popping 
> (PHP) for the LSP between PE1 and PE2.
> 
> 
>                             LSP
>                        ========>
> [CE1]------[PE1]-----------------[PE2]------[CE2]
> 
> 
> So, my question is who does push and who does pop here? Will top label "Y" 
> be ever used in this scenario?

The top label Y in this case is an implicit null (value=3) label. PE1 will
send a packet on the wire with only the one label in the label stack : label X.
Label Y is used in the signaling control plane by PE2 to indicate to PE1 to
do PHP for the LSP.

The label actions on PE1 are implemenation dependent. The simplest approach
is to regard this as a push X on PE1 as the encapsulation will be programmed
to push only one label.

> 
> Thanks for your response,
> Sachin Kalra
> 
> ===========================================================
> Sachin Kalra
> Modeling Engineer
> OPNET Technologies, Inc.
> (240)-497-3000 x2796
> ===========================================================
> Register for OPNET's Online Technology Workshops
> (http://www.opnet.com/TechWorkshops/)
> ===========================================================