The MPLS WG Archive

[Jean Philippe Vasseur <jvasseur@cisco.com>]

Hi Shahram,

At 06:21 23/08/2002 -0700, Shahram Davari wrote:
>Jean,
>
>I meant central.

right, the bypass LSP paths could be calculated:
- in a distributed manner by the PLRs
         o without bw protection (just easy, as the only constraint is to 
be diversely routed from the protected section),
         o with bandwidth protection (one method being described in 
draft-vasseur-mpls-backup-computation-00.txt)
- by a central PCS (one method being described in 
draft-vasseur-mpls-backup-computation-00.txt)

Thanks for your comments on both drafts.

JP.

>-Shahram
>
> > -----Original Message-----
> > From: Jean Philippe Vasseur [mailto:jvasseur@cisco.com]
> > Sent: Thursday, August 22, 2002 6:16 PM
> > To: Shahram Davari
> > Cc: 'Ping Pan'; 'mpls@UU.net'
> > Subject: RE: Comments on Fast Reroute draft
> >
> >
> > Hi Sharam,
> >
> > Sorry for the late reply ....
> >
> > At 11:16 09/08/2002 -0700, Shahram Davari wrote:
> > >Ping,
> > >
> > >draft-vasseur-mpls-backup-computation-00.txt is also proposing
> > >non-local bypass tunnel computation. Has this draft also
> > been rejected?
> >
> > could you clarify what was your original point regarding non
> > local bypass
> > tunnel path computation ?
> > JP.
> >
> > >-Shahram
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: Ping Pan [mailto:pingpan@cs.columbia.edu]
> > > > Sent: Friday, August 09, 2002 2:12 PM
> > > > To: Shahram Davari
> > > > Cc: 'mpls@UU.net'
> > > > Subject: Re: Comments on Fast Reroute draft
> > > >
> > > >
> > > > Shahram Davari wrote:
> > > >
> > > > > Ping,
> > > > >
> > > > >>The ingress does not need to compute the bypass tunnel.
> > > > Bypass tunnel
> > > > >>computation could be a local decision.
> > > > >>
> > > > >>http://www.ietf.org/internet-drafts/draft-ietf-mpls-fastrerout
> > > > >>
> > > > > e-mib-00.txt
> > > > >
> > > > >
> > > > > But it may also be a non-local decision.
> > > > >
> > > > > -Shahram
> > > > >
> > > >
> > > >
> > > > I think MPLS WG had gone through this argument before (there
> > > > was a draft
> > > > on this). I recall nobody wanted it. The question was why.
> > > > Also are we
> > > > going to design a RSVP fragmentation extension to support large
> > > > messages? ;-)
> > > >
> > > > - Ping
> > > >
> > > >
> > > >
> > > >
> >