Date: Fri, 28 May 1999 16:13:13 -0700 (PDT) From: Doug White <dwhite@resnet.uoregon.edu> To: Josef Belkovics <belkovic@osu.cz> Cc: freebsd-questions@FreeBSD.ORG Subject: Re: 3c509,ep,fbsd3.1 Message-ID: <Pine.BSF.4.03.9905281610550.11808-100000@resnet.uoregon.edu> In-Reply-To: <000d01bea91d$0bff9ec0$436a71c3@cl106067.osu.cz>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 28 May 1999, Josef Belkovics wrote:
> I am using fbsd 3.1 as ip+ipx router with 2x3c905 (xl,pci) + 4x3c509
> (ep,isa) boards.
Yuck. They do make quad-port PCI Ethernet cards. I don't know if a 66MHz
Pentium can handle those though.
> Ep boards many times stays in the state OACTIVE (tx is idle) which means
> that router does not route. I help myself with the following script:
>
> while sleep 5 ; do
> for i in ep0 ep1 ep2 ep3 ; do
> set X`ifconfig ${i} | grep OACTIVE`
> if [ X${2} != X ] ; then
> logger -t ifcon ${i}: ${2} (see syslog messages
> below)
> ifconfig ${i} down ; ifconfig ${i} up
> fi
> done
> done
>
Standard babysitter loop.
> I know that the same problem have various boards&drivers under linux. My
> 3c509 are connected in this way:
>
> May 26 05:01:16 se106012 /kernel: ep0 at 0x300-0x30f irq 7 on isa
> May 26 05:01:16 se106012 /kernel: ep0: aui/utp/bnc[*AUI*] address
> 00:20:af:b7:5
> May 26 05:01:16 se106012 /kernel: ep1 at 0x310-0x31f irq 10 on isa
> May 26 05:01:16 se106012 /kernel: ep1: aui/utp/bnc[*UTP*] address
> 00:20:af:e1:2
> May 26 05:01:16 se106012 /kernel: ep2 at 0x320-0x32f irq 11 on isa
> May 26 05:01:16 se106012 /kernel: ep2: aui/utp/bnc[*AUI*] address
> 00:a0:24:31:b
> May 26 05:01:16 se106012 /kernel: ep3 at 0x330-0x33f irq 12 on isa
> May 26 05:01:16 se106012 /kernel: ep3: aui/utp/bnc[*UTP*] address
> 00:a0:24:31:b
Interesting.
>
> ep0 => aui+fiber_optic_tranceiver => switch
Okay, this is the upset connection.
> ep1 => utp+fiber_optic_tranceiver => bridge (2x2Mb microwave connection)
> ep2 => aui+fiber_optic_tranceiver => switch
> ep3 => switch
>
> Router runs 4 years. Before installation of switches it routes well.
> Systat -vmstat 1 reports that p5/66 has enough capacity. I have raised
> timer tick (HZ=1000) but error has improved only little. The most little
> errors has ep1.
>
> Where is bug? In ep driver or in switches? (Please do not advise to connect
> 3c509 boards back into hubs - it is too late.)
>
> May 28 11:06:39 se106012 ifcon: ep0:
> flags=8c43<UP,BROADCAST,RUNNING,OACTIVE,SI
Perhaps the cabling between the swtich and ep0 is faulty? It looks like
the card's getting confused and is resetting itself.
The ep driver isn't the world's best.
Doug White
Internet: dwhite@resnet.uoregon.edu | FreeBSD: The Power to Serve
http://gladstone.uoregon.edu/~dwhite | www.freebsd.org
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.03.9905281610550.11808-100000>