On 1/19/2006 11:37 PM Rumen Yotov said the following:
On (19/01/06 18:15), Drew Tomlinson wrote:
On 12/29/2005 3:33 PM Drew Tomlinson wrote:
I have a AMD Athlon 2800+ processor running on a Asus A7N8X Deluxe
motherboard. This motherboard uses the Nvidia NForce2 chipset. It
has both an Nvidia and a 3Com 100 mbps network adapter integrated.
Initially I used the Nvidia network adapter but noticed a lot of Rx
errors. The "errors" and the "frame errors" matched exactly. I did
the standard testing to try and isolate the problem. I tried known
working patch cables and known working hub ports. Then when the
errors persisted, I even moved the box to a known working cable run.
The errors remained.
So I figured that the actual port on the motherboard must be bad. I
recompiled my kernel with 3Com networking support and began using the
3Com port but I am still getting errors. Here's my current 'ifconfig'
output:
eth2 Link encap:Ethernet HWaddr 00:26:54:0C:60:8D inet
addr:192.168.1.6 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:131785 errors:481 dropped:0 overruns:0 frame:760
TX packets:168955 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:27189676 (25.9 Mb) TX bytes:169280962 (161.4 Mb)
Interrupt:18 Base address:0xb000
The errors have changed a little in that now the "frame errors" exceed
the "errors". How can that be? I must not understand the output.
Anyway, I'm starting to think I may have some configuration issues as
I find it hard to believe that both ports are bad. I'm a Gentoo
newbie and started with kernel 2.11. I've upgraded once to 2.13-r5
which is where I'm at now. I have not changed any networking
parameter from their defaults as far as I recall. Any ideas on what
might be going on?
I'm still having this issue. A networking guy told me that framing
errors can typically come when the two ends of the connection (card and
switch) are not set to the same parameters such as half-duplex vs.
full-duplex. He suggested I try forcing both ends to be the same and
see if the errors go away.
However this is just a cheap 8 port TrendNet switch that doesn't have
any management capability. Thus I thought I'd try forcing the NIC to
one way or the other and see if the errors are resolved. I've Googled
for information on how to (a) find out the current setting and (b)
change the setting but have been unsuccessful. Can any one point me in
the right direction?
Thanks,
Drew
Hi,
For setting up NICs check: mii-diag or better ethtool (both are in portage).
Read the corresponding man pages for needed options.
HTH.Rumen
Thank you. I installed ethtool. The output of concern is this:
tv mythtv # ethtool -S eth1
NIC statistics:
tx_deferred: 0
tx_multiple_collisions: 0
rx_bad_ssd: 47853
Does rx_bad_ssd mean "framing errors"? I Googled on the term and only
came up with a bunch of ethtool patches.
Here's some other output if that helps anyone know why I might be
getting framing errors:
tv mythtv # ethtool eth1
Settings for eth1:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Advertised auto-negotiation: Yes
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 2
Transceiver: internal
Auto-negotiation: on
Current message level: 0x00000001 (1)
Link detected: yes
tv mythtv # ethtool -i eth1
driver: 3c59x
version: LK1.1.19
firmware-version:
bus-info: 0000:02:01.0
Thanks,
Drew
--
gentoo-user@gentoo.org mailing list
