Date: Mon, 16 Oct 1995 15:35:37 -0400 (EDT) From: Mark Hittinger <bugs@ns1.win.net> To: questions@freebsd.org Subject: Re: Too many UUCP checksum errors! Help! (fwd) Message-ID: <199510161935.PAA05076@ns1.win.net>
next in thread | raw e-mail | index | archive | help
> To: raoul@cssc-syd.tansu.com.au (Raoul Golan)
> > I've got one more problem I need help with (hopefully the last).
> > I've set up UUCP to a linux machine, and I'm finding that uucico
> > fails miserably due to too many checksum errors.
> > I've tried protocols i and g, and they both suffer from the same
> > problem.
We do a great deal of UUCP here and have just completed a project to
use large packet sizes with national ISDN access for our WinNET windows
uucp front end. FreeBSD delivers performance very close to the
theoretical maximums, even with all of our modem lights flashing.
(I know most people are not excited by UUCP, but it is a truely awesome
sight if you've been doing it using other supposedly "high performance"
[more expensive] solutions)
Nearly all of our uucp traffic goes into our FreeBSD servers via rlogin
from annex and portmaster boxes.
Here are some random things to check:
1. Sometimes people will have a motherboard com1/2 that is not disabled
when they have an additional com card with 16550's on it.
2. IRQ sharing can sometimes cause this.
3. Not enough swap space or other system resource wall issues can cause
this.
4. A cable that does not carry the hardware cts/rts flow control can
cause this.
5. Some other high interrupt activity on the ISA bus can cause this.
6. NEC 386sx-16 motherboards have an IRQ flaw which causes this.
7. We have had problems on VESA local bus motherboards where the serial
i/o is on a VESA card. The clock rate of the bus is above 33 MHZ.
If you have an AMD 486dx2/80 and your com ports are on a VESA local
bus IDE controller you will see this problem.
Regards,
Mark Hittinger
Internet Manager
WinNET Communications, Inc.
bugs@win.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199510161935.PAA05076>