The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2004-Mar> msg00069



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

draft-minei-mpls-ldp-planned-restart-00.txt

  • From: "Adrian Farrel" <adrian@olddog.co.uk>
  • Date: Thu, 18 Mar 2004 10:12:37 -0000
  • Cc: "Eric Gray" <ewgray@graiymage.com>, "'MPLS wg'" <mpls@UU.NET>


> > While this may be understandably frustrating to the authors of 3479 who foresaw this
need
> > (and many other needs) up front, The convergence of functionality is to be welcomed.
> >
> I am glad to hear you find the proposal useful.

Did I say that? :-)

> > Ina, are you sure that the function is needed only on a per-session
> > basis and that you will not want to discriminate recovery per-LSP?
>
> Currently I would like to keep this at a per-session level. If I
> want graceful, seamless behaviour, I want it for all my lsps, not just
> for some of them :)

Well this is your choice as an implementor. I am just opening up the question to the list.
I have no strong feelings on the issue.

Since not all labels are advertised for FECs that are installed from the same source (e.g.
some may be for static and local routes, some may be from different IGPs, some may be from
other routing sources) it might be conceivable that local policy wants to protect and
recover certain LSPs but discard others.

Cheers,
Adrian