RE: SDH/SONET in GMPLS

[James Scott <jscott@calient.net>]

hi manoj,

mj> In SONET, each VTG of STS-1 can have VTs of 
mj> different size (i.e 1 VTG can contain 1  VT-6, 
mj> other VTG within the same STS-1 can contain 2 
mj> VT-3 etc). As VTG in SONET is equivalent to 
mj> TUG-2 in SDH. Does this mean that each TUG-2 
mj> within STM-1 can contain different type of 
mj> VC-X (X=2, 12, 11) i.e. one TUG-2 can contain 
mj> 3 VC-12 and other TUG-2 within same STM-1 can 
mj> contain 1 VC-2 ?

yes.  a TUG-2 in SDH is equivalent to a VT group
in SONET.

mj> Furthermore, As per draft-mannie-mpls-sdh-ospf-
mj> isis-01.txt, It is assumed that every STM-1/STS-1 
mj> of STM-N/STS-M, M, N > 1 will have the same 
mj> multiplexing capability. This type of restriction 
mj> is not as per SDH/SONEYT. Does this mean this 
mj> restriction has been kept for making the routing 
mj> protocol simple ?

you need identical multiplexing capabilities at each
end of the LSP, in order to be able to demultiplex
at the LSP termination point.  the reason for
advertising this in the link state protocol is
to ensure consistency at each endpoint.  in other
words, by making this a CSPF constraint, we can
easily prune disjoint multiplex types from the
set of acceptable paths.

if this is unclear, perhaps i'll defer to the
authors for further clarification.  i hope that
my comments have been of use.

cheers,

james