Any way to force...

Cy llucius@tiny.net
Sat Apr 1 18:42:00 2000


...which physical interface gets assigned to which interface name?  I'm
pretty sure this was discussed recently, but did I play attention?  B-)  I
currently have all drivers compiled into the kernel, but am willing to
"modify" them if need be.  (Cool double usage there.  B-))  Just looking for
an easy way out.

Thanks much,

Leland

-----Original Message-----
From: owner-linux-tulip@beowulf.org
[mailto:owner-linux-tulip@beowulf.org]On Behalf Of Donald Becker
Sent: Saturday, April 01, 2000 10:25 AM
To: linux-tulip@beowulf.org; Metod Kozelj
Subject: Re: Never force full duplex.


On Sat, 1 Apr 2000, Metod Kozelj wrote:

> There's another consideration about FD vs. HD: if you have, say 4
> machines, all 100base-TX FD, and 3 of them want to push data to the fourth
> one full speed, the bandwidth to the fourth one is too narrow. Switch'
> buffers will become full and further packets from the first 3 machines
> will get silently dropped. High-level protocol will have to do the
> re-transmissions after timeouts.
>
> If all of them were HD, switch would insert collisions and low-level
> protocol would do re-transmissions, which is faster.

There is a relatively new feature, link flow control, that accomplishes the
same thing.  Many chips updated in the past two years have it.

It's negotiated in the same way as duplex, so forcing full duplex disables
it.  Even if you do remember to program your replacement switches with
forced-full-duplex, you are not getting the full potential of the hardware.


Donald Becker
Scyld Computing Corporation, becker@scyld.com

-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-tulip-request@beowulf.org

-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-tulip-request@beowulf.org