Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 04 Jan 2001 23:39:56 -0800 (PST)
From:      Nicole <nicole@unixgirl.com>
To:        freebsd-isp@freebsd.org, freebsd-net@freebsd.org
Subject:   Problem with fxp0 card and slowing/dying  transmits
Message-ID:  <XFMail.010104233956.nicole@unixgirl.com>

next in thread | raw e-mail | index | archive | help


 Hello all
 My poor hair is abt to come out as I keep pulling on it trying to solve what
is, to me, a Very Very weird problem.

 I have a server, running FreeBSD-3.5-STABLE as of 1/2/2000. It is a dual
400Mhz processor system with (I think) an Intel Motherboard (could also be a
tyan) that has built in (intel) Ethernet and SCSI. It has 256 Megs of Memory.
 It is connected to an Intel 460 switch along with abt 5 others servers very
similar to it.

 The problem is that when trying to scp a file or send a large file to it via
sendmail, (large = 253952 <a copy of sendmail>) it seems to transmit along
happily, then (at least for scp) at abt 1/2 way through, it seems to just start
crawling. When I have let it run, it will go forever and will seem to
eventually finish but will hang as the transfer speed keeps dropping through the
floor.

 I have tried numerous things, including shutting off the built in Ethernet
card and replacing it with a standard intel 10/100 pro card. (not the new ones
with the small VLSI chip, but the older style unit, exactly like what the other
systems have)

 I have tried altering net.inet.tcp.rfc1323 and net.inet.tcp.rfc1644. I have
tried setting the card into solid 100TX via ifconfig (ifconfig fxp0 inet
10.0.0.1 netmask 255.255.255.0 media 100baseTX ) with no effect.
 I even tried setting it to 10/BT with no improvement. Changed the port it was
in on the switch. Changed cables 3 times. Said several ancient prayers, and
even succomed to eating dead cow over it.

 The only other semi clue is that it was just moved from another ISP were it
was plugged into a Cisco switch and it seemed to be working fine there. All of
the other servers with the same card seem to work fine however via the same
Intel switch. You would think it would be happier, Intel card to Intel switch
anyway.

 ANY help or clues would be appreciated. Could this be caused by the MB? What
else can I try?

 Please CC me in any replies to make sure I see it right away.

   Thanks!!!


    Nicole






nicole@home:/home/nicole> sysctl -a | grep tcp
tcpcb:           288,     2344,    124,    142,     2761
net.inet.tcp.rfc1323: 0
net.inet.tcp.rfc1644: 1
net.inet.tcp.mssdflt: 512
net.inet.tcp.rttdflt: 3
net.inet.tcp.keepidle: 14400
net.inet.tcp.keepintvl: 150
net.inet.tcp.sendspace: 16384
net.inet.tcp.recvspace: 16384
net.inet.tcp.keepinit: 150
net.inet.tcp.log_in_vain: 0
net.inet.tcp.delayed_ack: 1
net.inet.tcp.pcbcount: 124
net.inet.tcp.always_keepalive: 1



 nicole@unixgirl.com         |\ __ /|   (`\   http://www.unixgirl.com/
 webmistress@dangermouse.org | o_o  |__  ) )  http://www.dangermouse.org/
 nicole@deviantimages.com   //      \\        http://www.deviantimages.com/     
 
---------------------------(((---(((----------------------------------------
 
              --  Powered by Coka-Cola and FreeBSD  --
  -- I don't speak for anybody but myself - that's enough trouble --
    --  Back Up My Hard Drive? I Can't Find The Reverse Switch! --

-------------------------------------------------------------------------
 


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.010104233956.nicole>