[c-nsp] Performance Degradation Using VLANs

[Pete Templin]

Rodney Dunn wrote:
> The performance hit would be very very little.
> 
> That's because the mac header rewrite is prebuilt
> by CEF to include the vlan/dmac/smac information
> and it's preprened just like if the vlan isn't
> there.
> 
> You would have to test it to get exact numbers but
> I've never seen anyone mention a difference in
> the software forwarding path when adding subinterfaces
> as a trunk that map to vlans.

Just for the sake of mentioning, this would be for 12.1(3)T (and later), 
12.2, 12.2T, 12.3, and 12.3T release trains.  I learned the hard way 
that subinterfaces in 12.0 do not seem to have full CEF support.

pt