On Thu, 18 May 2000, Matias Freytes wrote:
> Thank you once again.
> I remember I once used nediag or something similar by D.Becker.
> Will this be of any help? Where can I get it from?
http://www.scyld.com/diag/index.html
But that probably won't help here.
> > Look at the error statistics on the ethernet devices
> > ifconfig eth0
> > If there are error counts there interpert...
You should get the real error counts from /proc/net/dev.
> I'm sorry, I don't know how to interpret this. Here it goes.
> This is what I get on PC1 after some pinging to PC4:
>
> eth1 Link encap:Ethernet HWaddr 00:80:AD:19:55:4D
> RX packets:634 errors:0 dropped:0 overruns:0 frame:1
> TX packets:203 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:100
>
> This is what I get on PC4 after pinging PC1. No receiving errors but very
> high Tx error rate and collisions, and not usual base address:
>
> eth0 Link encap:Ethernet HWaddr 00:C0:DF:F3:BC:C4
> inet addr:192.168.1.4 Bcast:192.168.1.255 Mask:255.255.255.0
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:130 errors:0 dropped:0 overruns:0 frame:0
> TX packets:135 errors:117 dropped:0 overruns:0 carrier:117
> collisions:1989 txqueuelen:100
Baaadddd.
> Interrupt:10 Base address:0xe800
This is a PCI card. The base address will always be high to avoid ISA and
EISA cards.
> > >>If you can get a ping at all, your software is working. Sounds like you
> > >>have some dodgy ethernet wiring and perhaps poor connections.
...
> > >>> PCs 2 and 4 get aprox 95% loss when pinging enybody, even
> > >>> themselves, even with just the T and two line terminators
> > >>> (50 ohm).
Oh, 10base2.
OK, this is easy -- blame the cable. I used to be a big fan of coax when
10baseT was new and repeaters were expensive, unreliable active hardware.
But that was because I was always in facilities that had experts making the
cables and configuring the network. ("Ethernet is so much easier to work
with than Chaosnet hardline.")
Today everyone should be using twisted pair. You never have the problem of
"#1 can only talk to #4, #7 and #8".
Donald Becker [EMAIL PROTECTED]
Scyld Computing Corporation
410 Severn Ave. Suite 210
Annapolis MD 21403
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
