[tulip] Terrible throughput of KTI 221TX NIC (Intel 21143 chip) in fdx mode

Donald Becker becker@scyld.com
Mon, 26 Feb 2001 14:08:53 -0500 (EST)


On Mon, 26 Feb 2001, [iso-8859-1] Tamás Pusztai wrote:

>   I have a beouwulf type cluster of workstations.  The nodes
> have 3com 905c and KTI 221TX/3 NICs which are connected with
> a KTI switch.
...
>   However, the KTI cards show a sudden drop of the bandwith
> from 60Mb/s to 1-2Mb/s at around the blocksize of 5k in full
> duplex mode.  Half duplex is OK.

This is a typical duplex problem.

>   More details: I used the tulip.o driver supplied with RH7.0,
> more correctly, with RH7.0's kernel upgrade to 2.2.18. I also
> tried the de4x5.o driver, with similar results.  Is there any
> significant difference between these two drivers?

Please try the updated driver.  I'm confident that it will resolve your
duplex problem.
(I'm guessing that you have a 21143-TD chip with a SYM transceiver.)

>   And a last question:  What is flow control?  My switch has
> this parameter for each port, which seems to be enabled if the
> port is in autonegotiation mode and disabled if in forced modes.

Flow control is a unique multicast packet that is generated and
interpreted by the hardware.  It causes the transmitter to pause
transmits for a specified period.  (It must do this within one packet,
and thus usually must be interpreted by the hardware.)

It is negotiated, and thus is only valid on an autonegotiated link.

The 905C supports flow control.
Most tulip chips do not -- only the most recent chips have this feature.


Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Second Generation Beowulf Clusters
Annapolis MD 21403			410-990-9993