Re: The file that kills linux tcp

Arnt Gulbrandsen (
04 Feb 1997 21:49:33 +0100 (Alan Cox)
> > In that case, it would depend on the IP address of both the sending
> > and the receiving ends... which is just tricky enough that it could be
> > true.
> No you only need about 200 bytes of the right pattern to derail faulty
> 100baseT.

Since I get constistent errors with some clients (not distinguished by
OS, connectivity or hardware as far as I can see) and never errors
with others, I'd expect some destination addresses to form part of
these N bytes. But the source address doesn't matter - I've now tried
four different source addresses in two different /16 nets.

Btw, changing the card, cable or hub didn't help.