The MPLS WG Archive

[Hans Sjostrand <hans@ipunplugged.com>]

MPLS forwarding could be done by a device unable to do much else, in 
fact that was a few of the first implementations was made and it's also 
I think how a few of the GMPLS implementations today are done. An OXC
shouldn't have to terminate any wavelenght, just switch them. 

You use GSMP to separate the MPLS userplane (forwarding) from the 
MPLS control plane (signalling). Now you could MPLS (or GMPLS) enable
your old (or new) switch without putting in neither IP or MPLS protocools
in it. You just add a GSMP agent to the switch. You let the GSMP controller
do all the protocool stuff. 

Check out the GSMP WG if this interests you. 

Regards
/// Hasse 

> -----Original Message-----
> From: owner-mpls@UU.NET [mailto:owner-mpls@UU.NET]On Behalf Of Torsten
> Mueller
> Sent: den 5 juli 2001 15:56
> To: mpls@UU.NET
> Subject: Do MPLS core devices have to implement IP?
> 
> 
> 
> 
> 
> Do MPLS core devices have to implement IP?
> RFC 3031 states:
>  -  MPLS forwarding can be done by switches which are capable of
>          doing label lookup and replacement, but are either not capable
>          of analyzing the network layer headers, or are not capable of
>          analyzing the network layer headers at adequate speed.
> 
> While this means that MPLS nodes do not have to implement IP at the
> forwarding plane, do these switches have to implement IP at the control
> plane
> (and the standard label distribution protocols)?
> Or is it also possible to distribute the labels by management procedures
> to be compliant to the MPLS standards?
> 
> 
> Thanks,
> Torsten
>