On 5 May 2000, Jes Sorensen wrote:
> >>>>> "shane" == shane <shane@bratnet.net> writes:
>
> shane> How much could the CPU load affect the numbers, most of the
> shane> tests were run on the machines while theye were activly doing
> shane> other stuff. But while on the private ethernet segments the
> shane> ethernets wen't being used by any other process, but the
> shane> machines were busy doing other processing? This is a
> shane> PentiumPro200 System, any idea how much ethernet load it should
> shane> be able to take?
>
> I remember 11MB/sec TCP on Fast Ethernet taking around 40% CPU on a
> 233MHz P2, these numbers are waaaay old though.
In other words, running the link flat out (one way) should take roughly
half the CPU power of Shane's box. It should be able to handle it,
provided it isn't doing anything else too strenuous.
I'd try to run these tests with the two machines in single user mode,
linkes directly via cross-over cables, and isolated from the rest of the
world. That should eliminate most external influences.
Next, I'd try a different NIC (different manufacturer and chipset, not
just another of the same type), to rule out NIC driver problems. (And a
different cable.)
I'd also try the experiment talking to a different type of machine - say,
FreeBSD or Linux.
Then, I might try a few different kernel versions (maybe a 2.0, 2.2 and
2.3) to see if that affects results.
At least one of these should give some results - can you try them please,
Shane??
James.
(KDE #17)
-
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/
This archive was generated by hypermail 2b29 : Sun May 07 2000 - 21:00:17 EST