[c-nsp] channelized OC-12 for aggregation

[Rob Shakir]

Hi Geoff,

A couple of comments - we're currently running a bunch of 7600 with a bunch of chSTM-1 SPAs in SIP-200 for n*DS0 -> DS-3 aggregation, so hopefully some of our experiences might be of interest.

We've actually found the 7600 with SIPs to be really reliable boxes, after a few issues in early SRC, with newer software, we haven't had any problems with these boxes. We're doing HDLC, some frame-relay legacy support, and MLPPP across E1 or STM-1 bearers. We've got a set of issues with the way that some alarms are parsed, which I've explained below, but in general we're pretty happy with these 7600s! It's worth mentioning too that the SIP-200 is relatively good at providing some relatively useful features, so for example, doing LLQ inside a customer E1 is something that we've standardised on this platform.

I'm not sure that I can provide a good comparison against the ASR, but I don't think that the GSR option sounds cost effective in general. You will probably be aware that the 12000 series GSRs are EoS, which means that you will find that getting upgraded code (if you bought used hardware) is not possible. In addition, it is worth checking that 12.0(32)S(Y)? or 12.0(33)S support the features that you need.

On 13 May 2010, at 17:12, Geoffrey Pendery wrote:
> So my real questions :
> 1.  Anybody have experience with this particular SPA?  Good/bad?

Whilst not specifically this SPA - but the choc3/stm-1 variant of it, at least in the SDH world, there are ongoing issues with the parsing of the POH for a VC-12 channel. Cisco did not implement full parsing of the V5 byte, so therefore they are able to recognise only AIS for a specific VC-12. This is somewhat inconvenient when your provider doesn't send AIS on unequipped channels, and hence (correctly) sends LPUNEQ alarms. 

This is really inconvenient for us, since it causes issues with parsing of alarms that were available on older CHSTM-1 hardware. I am not sure whether this would have been the same for the choc-12/stm-4 card, but it might be worth checking if this is something that you're going to rely on.

> 2.  If we went with 7600, we'd have to switch to SRD from SXH - what
> sort of nasty surprises should I expect there?

We're much happier with SRD than we are with SXH as a train. Our target release for 7600s at the moment is SRD4, and we haven't found anything that's forced us to move away from it (yet!).

Kind regards,
Rob

-- 
Rob Shakir                      <rjs at eng.gxn.net>
Network Development Engineer    GX Networks/Vialtus Solutions
ddi: +44208 587 6077            mob: +44797 155 4098
pgp: 0xc07e6deb                 nic-hdl: RJS-RIPE

This email is subject to: http://www.vialtus.com/disclaimer.html