The MPLS WG Archive

[hai wang <yellowfall@163.net>]

Kamesh:

  I agree with you if we implement MPLS without considering comformance issues. For a RFC designed specifically for control-driven, data-driven implementation need to be checked carefully to avoid interconnection problem. :)

  Hai Wang

 

Hi:

       I think the question of whether MPLS LSP establishment is data or conrtol driven is an implementaion issue . The LSPs (static and through LDP) could be established apriori and the incoming traffic at the ingress can be mapped into them through policies .

Kamesh.

 

Kameswara Rao Avasarala

Software Engineer

Ericsson Inc

Telephone:

Work:805-562-6212

Resi: 805-562-8841

 

 

-----Original Message-----
From: owner-mpls@UU.NET [mailto:owner-mpls@UU.NET]On Behalf Of abhay
Sent: Wednesday, December 05, 2001 8:21 PM
To: hai wang; mpls
Subject: Re: Does MPLS Now specifically for control-driven?

Hai,

     Yes,I share the same thoughts as you do,

      also,consider a case wherein,on detection of flow's you can request the control plane for a mpls job

      However,there have been only customized tag or label protocols which are specially made for proprietary

      implementations.

      However data-driven is dependent on the control plane protocols for any mpls job.

      Regards,

      Abhay

----- Original Message -----

From: hai wang

To: mpls

Sent: Wednesday, December 05, 2001 9:16 PM

Subject: Does MPLS Now specifically for control-driven?

Hello everyone:
   I have read MPLS drafts, from drafts it seems that MPLS is not specifically designed for data-driven or control-driven(although one may think control-driven is better), which I think that it allows either method to be used in label distribution. However, after reading RFC 3031, I got a impression that MPLS now is specifically designed for control-driven method, for it is binded to routing protocols tightly.
   Is it a change in standard track? or is there anything I am misunderstanding?
   Thanks.
Hai Wang.

---