[c-nsp] Curious... 2800 or 3800

[Lincoln Dale (ltd)]

> Assuming that the CEF entry had already been made and that there
wasn't
> going to be any incurred overhead from processing switching the first
> packet.

that's the whole point of CEF.  there is no "process switch the first
packet"; think of CEF as being the table is ALWAYS prepopulated; if it
isn't populated, the route doesn't exist. :)

> but I would say that it is highly unlikely that
> high-overhead transport mediums and ATM and POS were used for the
test.

actually, transport has little to do with it; SAR on ATM, bitstuffing on
POS etc. are handling by the framer (hardware) not software.


cheers,

lincoln.