Slow transfer with 21140 card

[Robert G. Brown]

On Wed, 22 Jul 1998, Mark Hahn wrote:

> > I just found that UDP appears to be transmitting correctly, whereas
> > TCP is slow.  Thanks to Mark Hahn for the suggestion.  Now what
> > does this mean and what do I do?
> 
> when this happened to me, the cause was somewhat embarassing.
> I had wired the cables/connectors myself, using good cat5,
> not abusing the cable, testing the crimp, etc.  unfortunately,
> I had assumed that RJ45's were wired with pairs on adjacent pins
> (ie, orange-white,orange,grn-wh,grn,blu-wh,blu,brn-w,brn).
> unfortunately, it's not sane like this: the green pair is split,
> and the reversed blue pair is in between.  the result was that my
> Tx signal pair (I think) was now on wires in two pairs, that is,
> not twisted.  obviously, this is very lossy, and I was surprised
> that it ever worked.  note also that it worked pretty well at 
> 10bT, as well...
> 
> anyway, once I corrected this, all was well.  I can actually see
> ~9 MB/s using TCP, even with HD on a hub.  UDP runs at 11+ MB/s
> in either direction...

This makes sense, as UDP doesn't do much error testing while TCP will
force a retransmit of bad packets.  Bad connectors is one possible
source of corruption, but there are others.  Something to pursue.

  rgb

Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb@phy.duke.edu