RE: RE: High Speed Firewalls

["David Newman" <dnewman () networktest com>]

> > From the application layer perspective, I believe that 12.5 Mbyte is
> the theoretical maximum.  Practical throughput is, about, 80% of
> max...
>
> Regards,
> Gregory Hicks

There's a reason the theoretical limit is higher than the practical
one--there's overhead involved. That includes stuff like packet headers and
tcp and ftp setup and teardown.

Even on a perfect network -- a piece of wire between two hosts with zero
collisions -- I still wouldn't be able to transfer a 12.5-Mbyte file in 1
second over a fast Ethernet link. I can probably do much better than 80
percent -- I should be able to get well into the 90s -- but I won't hit the
theoretical max if I'm measuring from an application-layer perspective.

This brings me back to my original point -- when benchmarking firewall
performance it's important to be mindful of what we're measuring, and from
what perspective.

David Newman