Re: GRSVP-TE and MSSPRING

["Adrian Farrel" <adrian@olddog.co.uk>]

Yes, that covers it.
It is a sort of crankback.
Adrian
----- Original Message ----- 
From: "Diego Caviglia" <Diego.Caviglia@marconi.com>
To: "Adrian Farrel" <adrian@olddog.co.uk>
Cc: <ccamp@ops.ietf.org>
Sent: Wednesday, November 26, 2003 11:25 AM
Subject: Re: GRSVP-TE and MSSPRING


>
> Hi Adrian,
>           some doubts about the solution you suggested arises in my mind.
>
> Suppose we are in the following scenario.
>
>                                      |                   MSSPRING
> |
> TNE1<------>TNE2<------>TNE3<------>TNE4<------>TNE5<------>TNE6<------>TNE7<------>TNE8
>                                             1          1            1
> x
>                                             2          2            2
>  2
>
> TNEs 4, 5, 6, 7 and 8 are part of a MSSPRING.
> Number below the Link indicates the Label availability an x indicates that
> the Label is not available.
>
> Up to TNE 4 the signallin procedures as usual then TNE4 has to select the
> UPSTREAM LABEL and it choose Label1.
>
> Label1 is good up to TNE 7 that refuses it (it refuses the Label1 because
> it knows that is part of a MSSPRING and thus cannot change the Label,
> otherwise for "normal" -"normal" menas no MSSPRING- signalling an UPSTREAM
> LABEL with value 1 is good for TNE7) and issues a PathErr with
> AcceptableLabelSet that indicates Label2.
>
> Now my concern is about who process the PathErr with the
> AcceptableLabelSet.  Is TNE6 is TNE1 or is TNE4?
>
> As far as I've understood is TNE 6.
>
> TNE6 removes the XC, if it created it, and propagates the PathErr Upstream.
>
> The Upstream propagation of the PathErr with the AccetableLabelSet is
> fundamental for the procedure.
>
> TNE5 behave as TNE6.  TNE4 now have the right information and can send out
> a Path Message with Label2.  Is my interpretation correct?
>
> If yes it seems a sort of crankback.
>
> Regards.
>
> Diego
>
>
>
>
> "Adrian Farrel" <adrian@olddog.co.uk> on 25/11/2003 22.44.33
>
> Please respond to "Adrian Farrel" <adrian@olddog.co.uk>
>
> To:    <ccamp@ops.ietf.org>, "Diego Caviglia" <Diego.Caviglia@marconi.com>
> cc:
>
> Subject:    Re: GRSVP-TE and MSSPRING
>
>
> Diego,
>
> I believe that if you use Label Set and Acceptable Label Set you will
> arrive at the right
> result.
> It may take a few messages backwards and forwards to sort it out.
>
> Regards,
> Adrian
> ----- Original Message -----
> From: "Diego Caviglia" <Diego.Caviglia@marconi.com>
> To: <ccamp@ops.ietf.org>
> Sent: Tuesday, November 25, 2003 3:00 PM
> Subject: GRSVP-TE and MSSPRING
>
>
> > Hi all,
> >               MSSPRING has been defined in ITU-T G.841 and is widely used
> > as inherent protection scheme in transport network (SDH).
> >
> > MSSPRING introduces some constraints to Label selection due to the
> > impossibility to change TimeSlot (a.k.a. Label) inside a MSSPRING ring.
> >
> > In case of bi-directional LSP Upstream TNE has to choose the Label and
> > signal it by means of the UPSTREAM Label obj but if the TNE is inside a
> > MSSPRING ring it MUST choose a Label that is free on all the TNE
> traversed
> > by the LSP that are part of the MSSPRING ring.
> >
> > Of course the problem is that it does not have that information.
> >
> > My question is how can I signal, with GRSVP-TE, a bidirectional LSP that
> > traverses an MSSPRING ring?
> >
> > My impression is that I can not.
> >
> > Any answer is welcome.
> >
> > Regards
> >
> > Diego
> >
> >
> >
> > ------------------------------------------
> > Diego Caviglia
> > Marconi - Optical Networks
> > ASTN/GMPLS - System Design Manager
> > Via A. Negrone 1/A
> > 16153 Genoa - Italy
> > Phone: +390106003736
> >
> >
> >
> >
> >
> >
>
>
>
>
>
>
>
>
>
>
>