errors with ifconfig

[ismike]

I'm running Slackware 7.1 with two 3Com905B network cards. eth0 connects to my cable modem and eth1 connects to my LAN. When using my 10Base-T (generic) hub I have no problems, but recently I bought a NetGear fast ethernet switch (FS108). Now I get errors on RX packets on eth1. Also when pinging it from other machines it will time out once in a while (about 48% packet loss). Any ideas b/c this has me completely stumped. Usually I'm pretty self-sufficient, but I've been stumped on this for a while. The other computers can ping one another just fine, but none of them can ping the computer in question without problems. I'm also including sample output.

#ifconfig
eth 1 Link encap:Ethernet HWaddr 00:01:02:6B:82:11
inet addr: 192.168.0.1 Bcast:192.168.0.255 Mask 255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:49 errors:18 dropped:0 overruns:0 frame:17
TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen: 100
Interrupt:11 Base address:0xe800

Any help that any of you can provide would be greatly appreciatet. Thanks in advance. :confused: null :confused: :confused:

[jumpedintothefire]

Ok this might sound nuts, but give it a reboot.

I've had a bit off a hard time doing a hotswap with that card. :eek:


Even after a (I use RH)
/sbin/ifdown eth1
/sbin/ifup eth1
:(

I could not get the full duplex light on the switch to come on when pluging in. :confused:
WTF??

I think it is because of nway card and switch not doing the auto setup. :mad:

Reboot the light came on :o

and all was well. :D


Try a different cable next, somtimes a cable that works ok a 10, might be marginal when used with 100. Try a diffent port on the
switch, might be a bad port from the factory.


Hope it helps...

[ 17 November 2001: Message edited by: jumpedintothefire ]

[ismike]

Thanks for the suggestion. I have rebooted several times and have tried several different ports on the switch. The other computers work fine in all of the ports :mad: I will try a different cable (I have one that works fine at 100 on another computer). I'm wondering if maybe the drivers for the network cards are compiled into the kernel instead of being loaded as modules? Would this be a problem if you had one driver compiled in but had two cards running at different speeds trying to use the same driver? This could be completely off base because this is where I still don't have a lot of knowledge about Linux. Let me know what you think and thanks again for all your time and help! :D

[jumpedintothefire]

try:

lsmod should list the modules loaded.
http://www.ibiblio.org/pub/Linux/docs/HOWTO/Ethernet-HOWTO

It has the best detail on the subject, and a bunch of stuff to check. Along with info on how to pass kernel options at boot time,

[jumpedintothefire]

http://www.scyld.com/mailman/listinfo/vortex-bug

This has the archives for the nic card driver, have a read.

[ismike]

Update: I'm currently reading through the information provided at the links that were listed.

I tried a different cable which I know works at 100Mbps and had the same results. Also doing an lsmod provides the following information:

Module Size Used by
ip_masq_icq 13272 0 (unused)
ip_masq_raudio 2904 0 (unused)
ip_masq_irc 1592 0 (unused)
ip_masq_ftp 2584 0
3c59x 19264 2
bsd_comp 3764 0 (unused)
ppp 20652 0 [bsd_comp]
slip 7388 0 (unused)
slhc 4504 0 [ppp slip]
lp 5924 0 (unused)
parport_pc 7396 1
parport 6952 1 [lp parport_pc]

I will continue to read and try to figure this out. I appreciate everyone's help thus far. Thanks! :D