[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Generalizing WSON information...
Greg,
The work on switching capabilities, RFC4202, assumes a pool of node
resources shared among links and was advertised as a link capability.
When a given pool of resources was exhausted, the corresponding link
capability was withdrawn for all links.
Thanks,
John
> -----Original Message-----
> From: Greg Bernstein [mailto:gregb@grotto-networking.com]
> Sent: Wednesday, April 15, 2009 10:02 AM
> To: julien.meuric@orange-ftgroup.com
> Cc: ccamp@ops.ietf.org
> Subject: Re: Generalizing WSON information...
>
> Hi Julien, interesting point. It did seem like older SDH
> equipment did have a "time-slot continuity constraint", but I
> never directly dealt with those systems. There was some work
> a while ago on MS-SPRing (Diego?).
> Maybe we need to think up a level, as you said in the usual
> GMPLS, MPLS cases full label swapping is a typical
> capability. Should we be explicit in the WSON case (maybe
> with the connectivity matrix) in indicating that this may or
> may not be the case? Right now we are being implicit about
> this lack of wavelength (label) conversion capability.
>
> The notion of the shared wavelength converter pools and the
> accompanying details seem very WSON specific. But the fact
> that a switch or mux can't perform label exchange seems a
> fairly general notion that we could group with other
> generalizable information...
>
> Best Regards
>
> Greg
>
> julien.meuric@orange-ftgroup.com wrote:
> > Hi Greg.
> >
> > Just a small comment on Wavelength Converter Pool in
> section 3.2. In some other GMPLS contexts, this may be
> modelled as a label swapping capability, which is a common
> enough to be considered as "Generalizable" (it's even a
> typical case that becomes an exception due to our label
> continuity constraint). However, as for several other
> parameters, I don't think there is an actual need for it
> (this is already the default); but maybe we could imagine
> some MS-SPRing operations in SONET/SDH.
> >
> > Regards,
> >
> > Julien
> >
> > -----Original Message-----
> > From: owner-ccamp@ops.ietf.org [mailto:owner-ccamp@ops.ietf.org] On
> > Behalf Of Greg Bernstein
> > Sent: Tuesday, April 14, 2009 6:32 PM
> > To: ccamp
> > Subject: Generalizing WSON information...
> >
> > Hello fellow CCAMPers, at the 74th IETF meeting in San
> Francisco the
> > question came up as to what if any of the WSON path computation
> > information model would be useful in other technologies. Below is a
> > first attempt at such an assesment based on the current version of
> > draft-ietf-ccamp-rwa-info-02.txt.
> > Existing GMPLS
> > standard information is not included.
> >
> > From section 3.2 Node Information:
> >
> > Switched Connectivity Matrix: Generalizable:Yes. This can
> be used to
> > model any type of asymetrical switch in any technology. Caveat:
> > Besides optical is there any current products that can make use of
> > this?
> >
> > Fixed Connectivity Matrix: Generalizable: Yes. This can be used to
> > model fixed connectivity between ports. Caveat: Is there any need
> > outside optical?
> >
> > Wavelength Converter Pool: Generally useful: No. This is very
> > application specific to optical switching systems.
> >
> > From section 3.3 Link Information:
> >
> > Switched and Fixed Port Wavelength (label) Restrictions: Generally
> > useful: I don't think so but open to examples. These
> constraints must
> > be shared in the WSON case for two reasons: (a) the wavelength
> > continuity constraint requires global label assignment, (b) WSON
> > devices present many different types of wavelength
> constraints. Note
> > that without requirement (a) then (b) doesn't need to be
> shared since
> > local label (wavelength) assignment would suffice.
> >
> > From section 3.4 Dynamic Link information
> >
> > Available Wavelengths (labels): Generally useful? We need detailed
> > wavelength availability information due to the wavelength
> continuity
> > constraint.
> > Way back
> > in the old days many of us implemented bit maps to track SONET/SDH
> > time slots but this never made it into the standards. Any
> interest in
> > that now? Other examples?
> >
> > Shared Backup Wavelengths (labels): Similar to the above
> but used in
> > efficient shared mesh backup path computation.
> >
> > From section 3.5 Dynamic Node Information
> >
> > Wavelength Converter Pool Status: Not generally applicable. Too
> > application specific.
> >
> > Comments appreciated
> >
> > Greg
> >
> >
>
> --
> ===================================================
> Dr Greg Bernstein, Grotto Networking (510) 573-2237
>
>
>
>