Matthew Wilcox wrote:
> On Mon, Jun 28, 1999 at 10:47:39AM +1000, Richard Gooch wrote:
> > Larry McVoy writes:
> > > The proposal is this:
> > >
> > > a) get VA or Penguin to sponsor the hardware and the test lab,
> > > I'll talk to them and try and make that happen.
> > > b) create a test lab with enough clients to generate the load
> > > with a setup like so:
> > >
> > > [ server ]
> > > |
> > > gbit
> > > |
> > > [ router ]
> > > / / \ \
> > > / / \ \
> > > / / \ \
> > > c c ... c c
> > >
> >
> > I'd also suggest that the server be set up with multiple 100 Mb cards
> > (aka the Mindcraft test).
>
> [i've taken the liberty of correcting the name, as you yourself pointed
> out later]
>
> Why do you suggest that? One gigabit network card is cheaper than 4
> hundred megabit cards and provides more bandwidth. One of the reasons
> the Mindcraft benchmarketing exercise was so poor.
If you run benchmark tests with GB Ethernet you loose any right to complain
about unrealistic benchmarks that others might create. A machine with 4 x 100MB
Ethernet cards is a common real world configuration - often more for redundancy
than actual throughput. How many people have GB Ethernet installed, or will do
so in the next 12 months?
Also tell me where I can get a GB Ethernet card for US$70. I can get 4 x 100Mbps
high performance Tulip compatible cards for that. Also where could I get a GB
hub for less than 4 times a 100MB hub?
Lets get real here
Steve
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/