Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 17 Sep 1999 11:41:18 +0200
From:      Palle Girgensohn <girgen@partitur.se>
To:        dg@root.com
Cc:        river <river@theriver.nu>, freebsd-isp@FreeBSD.ORG
Subject:   Re: fxp0: device timeout
Message-ID:  <37E20CBE.BFC69E55@partitur.se>
References:  <199909170143.SAA15206@implode.root.com>

next in thread | previous in thread | raw e-mail | index | archive | help
David Greenman wrote:
> 
>    "simplex" isn't a duplex indicator; it indicates whether the interface can
> hear its own broadcasts. Assuming that you meant half-duplex, then it's
> probably the switch that is at fault, especially if it is a Cisco which
> always seems to get autonegotiation wrong.
> 


oops... I've never bothered to check this. Here's the report from
ifconfig:

$ ifconfig fxp0
fxp0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        inet 193.10.208.65 netmask 0xffffff00 broadcast 193.10.208.255
        ether 00:a0:c9:ac:e7:75 
        media: autoselect (100baseTX) status: active
        supported media: autoselect 100baseTX <full-duplex> 100baseTX
10baseT/UTP <full-duplex> 10baseT/UTP


I guess I can set the switch to full-duplex, no auto-sensing, and
ifconfig the NIC with 'media full-duplex', right?

/Palle


> -DG
> 
> David Greenman
> Co-founder/Principal Architect, The FreeBSD Project - http://www.freebsd.org
> Creator of high-performance Internet servers - http://www.terasolutions.com
> Pave the road of life with opportunities.
> 
> >> It sounds as if they have the switch and the netcard config'd differently.
> >> I.E. switch half duplex and NIC full duplex or vice versa.  I have seen this
> >> before and it causes horrible network performance just like you are
> >> describing.  If the NIC is set to 100mb full duplex have them check the
> >> switch.  Also, fxp0's are notorious for not detecting full duplex, check the
> >> log files if it is set to auto-detect....
> >>
> >
> >Ahh... It is set to auto-detect, and they claim it initially didn't
> >detect full duplex, and that's why they set the switch to simplex (so
> >they say, anyway, and I they're right). The NIC is running at simplex
> >speed. I still believe the problems span from network congestion, since
> >it works fine at off office hours, but this explains why they couldn't
> >run at full duplex.
> >
> >I'll set it to full-duplex, no autosensing. Is it the driver or the NIC
> >not detecting?
> >
> >Thanks!
> >
> >/Palle
> >
> >> hope this helps
> >>
> >> -----Original Message-----
> >> From: Palle Girgensohn [mailto:girgen@partitur.se]
> >> Sent: Tuesday, September 14, 1999 7:54 PM
> >> To: freebsd-isp@freebsd.org
> >> Subject: fxp0: device timeout
> >>
> >> Hi!
> >>
> >> fxp0: device timeout... I suppose this probably means that the network
> >> is statured, right?
> >>
> >> Are there any problems running fxp0 in full duplex?
> >>
> >> This machine is running at a client's site, and they complain it's slow,
> >> still the CPU is idle and I get the impression their entire LAN is
> >> completely satured. They had problems starting the machine in full
> >> duplex so they set the NIC up in simplex (by configuring the network
> >> switch it is connected to). Can they just pull the rj45 LAN connection,
> >> reconfig the switch and put the cord back in, or do they need to do a
> >> ifconfig down && ifconfig up?
> >>
> >> FreeBSD-3-STABLE from beginning of July. 2 proc SMP, 400 MHz, lots of
> >> RAM.
> >>
> >> Thanks for any input.
> >>
> >> /Palle
> >>
> >> To Unsubscribe: send mail to majordomo@FreeBSD.org
> >> with "unsubscribe freebsd-isp" in the body of the message
> >>
> >> To Unsubscribe: send mail to majordomo@FreeBSD.org
> >> with "unsubscribe freebsd-isp" in the body of the message
> >
> >
> >To Unsubscribe: send mail to majordomo@FreeBSD.org
> >with "unsubscribe freebsd-isp" in the body of the message


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?37E20CBE.BFC69E55>