[c-nsp] Maximum spannig tree instances

[A.L.M.Buxey at lboro.ac.uk]

Hi,

> purchasing.  The docs seem to clearly state that the limits are per-slot and
> do not mention model numbers. However, I can confirm that I have greatly
> exceeded this specification for years now without serious wonkyness.  I
> have WS-X6516A-GBIC cards running as high as 6,400 virtual port instances.
>  I do notice RSTP isn't quite as rapid as it used to be though.  If all STP
> instances reconverge at the same time, it might take a second or two.  If
> only one VLAN reconverges, it is still sub-second.

I have come across a couple of docs now that suggest that the 67xx are
not tied to this limit....but then they dont specify what the new limit
is and there must be one.. ;-) - nothing so clear for the 65xx line cards.
however, i came across more resources which suggest limiting and their
best practice docs also state you should limit what goes down the trunks.
suffice to say - rather than leaving the network 'exposed' to such
random limits and risks I feel its best to go along with more control
over the core<->aggregation layer.  (limiting right out to the distribution layer
seems to be a much larger resource step both for deployment and day to day ops)

alan