The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2003-Mar> msg00304



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

Distributing Labels for PWs within an LSP TE Tunnel

  • From: Shahram Davari <Shahram_Davari@pmc-sierra.com>
  • Date: Fri, 21 Mar 2003 09:06:29 -0800
  • Cc: mpls@UU.NET, "Mustapha Aissaoui (E-mail)" <mustapha.aissaoui@alcatel.com>

Alia,

I think Nurit's question was how can we bind a PW to a transport LSP, in a way that both PE's know about it? This would be useful if LDP uses per-interface label space.

An extension to PW signaling may be needed. Or a DoD label request that is pushed inside
the tunnel LSP could be used.

-Shahram



>-----Original Message-----
>From: Alia Atlas [mailto:aatlas@avici.com]
>Sent: Thursday, March 20, 2003 2:08 PM
>To: Nurit Sprecher
>Cc: mpls@UU.NET
>Subject: Re: Distributing Labels for PWs within an LSP TE Tunnel 
>
>
>At 11:42 AM 3/20/2003 +0200, Nurit Sprecher wrote:
>
>>Hi all,
>>
>>I am looking for a method for distributing PWs' labels of PWs 
>within a TE
>>LSP tunnel.
>>The TE tunnel is signaled using the RSVP-TE signaling 
>protocol. Is there a
>>method for distributing labels for PWs that should run within 
>TE LSP tunnel?
>>I am aware of the IETF draft for signaling the PWs' labels 
>using the LDP
>>protocol, but I would like to attach a certain PW with a 
>certain TE LSP
>>tunnel. How can I gain it? Maybe I should activate the LDP 
>for a certain PW
>>within the required tunnel (inband)? Is there another way to 
>do it but not a
>>static one (by configuration)?
>
>Hi,
>
>You can use a targeted LDP session to distribute the labels for a PW 
>between the two PEs.  How that PW travels across the PSN is 
>independent of 
>how the labels for the PW were distributed.  Thus, the PW 
>could use either 
>an LDP LSP or a RSVP-TE LSP to reach the egress PE.  Which one 
>to use and 
>how the PW is mapped to it is an internal issue on the ingress PE.
>
>Alia
>
>
>>I wonder also if there is a way to signal a TE PW?
>>
>>I'll appreciate your response.
>>Thanks in advance, Nurit.
>
>