[j-nsp] non-concatenated STM-4 on Junipers

[Lars Erik Gullerud]

First of all, let me say I don't really have too much personal
experience with SDH/SONET on a transmission level, so excuse me if this
question is stupid. :)

I'd like to provision an STM-4 over an existing STM-16 SDH-ring to
connect a couple of M40's in two POPs, since I happen to have some
1-port OC-12 PICs available and the capacity is currently unused on the
ring.

The problem is that the ADMs used on this ring can not add/drop a
concatenated STM-4, only non-concatenated (i.e. on the STM-4 ports it
can only drop 4 x VC-4, not a VC-4-4c). In fact, there are not tribs
available for this particular ADM platform who can do concatenation on
STM-4 level, so we'd have to replace the ADMs if we want to deliver
concatenated STM-4's on them. This currently seems to be a show-stopper,
since most router linecards don't seem to handle non-concatenated SDH
circuits.

However when I looked at the specs for the 1-port OC-12 PIC's, they say
"operates in both concatenated and non-concatenated mode", so I'm
suddenly seeing a tiny spark of hope. But does that mean I need to set
the "no-concatenation" sonet-option on the PIC and be presented with
four STM-1 channels, or will it actually do concatenation internally on
the PIC and present me with a full STM-4 on the Juniper?

If given four STM-1 channels, I suppose we could configure a "SONET
aggregated interface" in the software and re-bundle the channels
logically, but I'd obviously prefer to be presented with a single STM-4
interface. Is this possible without concatenation being done on the
ADMs?

Hopefully someone who is a bit more clueful about SDH (and the Juniper
SONET PICs) than myself know more about this?

/leg