The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2001-Aug> msg00241



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

Small comment on draft-ietf-mpls-te-feed-02.txt

  • From: "Don Fedyk" <dwfedyk@nortelnetworks.com>
  • Date: Tue, 21 Aug 2001 19:23:24 -0400
  • Cc: "Bilel Jamoussi" <jamoussi@nortelnetworks.com>, "Darek Skalecki" <dareks@nortelnetworks.com>, "'mpls@uu.net'" <mpls@UU.NET>
  • X-Orig: <dwfedyk@americasm01.nt.com>

Title: RE: Small comment on draft-ietf-mpls-te-feed-02.txt

Hi Adrian

This seems like a reasonable request. We can add text around this.

There is one point though, that the reporting zero bandwidth is
logically independent to the node going down or the removal of LSPs.
Actually, we use zero feedback to mean temporarily avoid this
link as a default case from situations like this. It is important to
remember the adjacent node sees the real bandwidth but due to LDP
being in an indeterminate state his option is to report the link
as being full (zero bandwidth). The restaging of the LDP adjacencies
should trigger corresponding link LSA to reaffirm the actual bandwidth
in these cases.

Regards,
Don

Adrian Farrel wrote:

Peter,

Thanks for this draft which looks stable to me.

Would it be possible to add a reference to LDP FT procedures (draft-ietf-ldp-ft) in section 5?  I think the third paragraph should have the following text appended...

  Where LDP FT procedures [LDP-FT] are in use, LSPs that are protected
  by FT procedures should not be torn down until after session re-
  establishment has failed.  Further, feedback TLVs should not report
  zero bandwidth availability on the link until after session re-
  establishment has failed.

Thanks,
Adrian
--
Adrian Farrel
Movaz Networks Inc.
afarrel@movaz.com