Network not fully functional
David Cato
2.2.14. Occassionally, the network will start correctly during a boot,
but about 90% of the time it does not.
Normally, the machine in question cannot ping other hosts on the network
although other hosts can ping it (sometimes the other hosts cannot ping
it either), and this usually causes the network to come up, but not
always.
This happens with 3 different cards (3Com 3c905C, Netgear FA310TX, and
Linksys NC100), so it looks like something more basic than a card or
driver issue. In each case, the cards are correctly detected during
system boot. For example, the 3Com results in the following output
during boot (no error messages are printed):
3c59x.c:v0.99H 11/17/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/
vortex.html
eth0: 3Com 3c905C Tornado at 0xd800, 00:50:04:c3:81:29, IRQ 18
Internal config register is 1800000, transceivers 0xa.
8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface.
MII transceiver found at address 24, status 782d.
Enabling bus-master transmits and whole-frame receives.
/proc/pci shows (why the unknown device?)
Bus 0, device 10, function 0:
Ethernet controller: 3Com Unknown device (rev 108).
Vendor id=10b7. Device id=9200.
Medium devsel. IRQ 18. Master Capable. Latency=32. Min Gnt=10.Max Lat=
10.
I/O at 0xd800 [0xd801].
Non-prefetchable 32 bit memory at 0xf9800000 [0xf9800000].
`ifconfig eth0' shows
eth0 Link encap:Ethernet HWaddr 00:50:04:C3:81:29
inet addr:192.168.42.100 Bcast:192.168.42.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1997 errors:0 dropped:0 overruns:0 frame:0
TX packets:1165 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
Interrupt:18 Base address:0xd800
and `route -n' shows
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
192.168.42.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo
0.0.0.0 192.168.42.1 0.0.0.0 UG 1 0 0 eth0
`tcpdump' generates the following:
[First, the client pings the gateway and gets no response...]
15:04:25.679224 arp who-has 192.168.42.1 tell 192.168.42.100
15:04:26.679280 arp who-has 192.168.42.1 tell 192.168.42.100
[Now, the gateway pings the client and gets a response...]
15:04:26.691177 arp who-has 192.168.42.100 tell 192.168.42.1
15:04:26.691620 192.168.42.100.1027 > 192.168.42.1.53: 5425+ A? ntp. (21)
15:04:26.691729 arp reply 192.168.42.100 is-at 0:50:4:c3:81:29
15:04:26.692174 192.168.42.1 > 192.168.42.100: icmp: echo request
15:04:26.692359 192.168.42.100 > 192.168.42.1: icmp: echo reply
15:04:27.076351 192.168.42.1.53 > 192.168.42.100.1027: 5425 NXDomain* 0/1/0 (94)
15:04:27.682062 192.168.42.1 > 192.168.42.100: icmp: echo request
15:04:27.682343 192.168.42.100 > 192.168.42.1: icmp: echo reply
.
. <trimmed several request/reply records>
.
15:04:31.689564 arp who-has 192.168.42.1 tell 192.168.42.100
15:04:31.689968 arp reply 192.168.42.1 is-at 0:20:78:e0:5b:c1
.
. <trimmed some more request/reply records>
.
[Finally, the client gets a response from the gateway...]
15:04:43.957985 192.168.42.100 > 192.168.42.1: icmp: echo request
15:04:43.958620 192.168.42.1 > 192.168.42.100: icmp: echo reply
Unfortunately, this doesn't always last and eventually the client can
no longer talk with the rest of the network.
I had things working perfectly for over 2 months with the Linksys card
originally until the system was taken down to add a second SCSI card.
However, the second SCSI card doesn't seem to be at issue as the problem
still exists if I remove the SCSI card. I then tried the 2 other
ethernet cards with no change in behavior.
Can anyone point me in the right direction to fix this problem?
--
dcato AT crunchyfrog DOT net
No Microsoft programs were used in the
creation or distribution of this message.
Vilmos Soti
> I'm having problems getting network cards fully functional under Linux
> 2.2.14. Occassionally, the network will start correctly during a boot,
> but about 90% of the time it does not.
...
> This happens with 3 different cards (3Com 3c905C, Netgear FA310TX, and
> Linksys NC100), so it looks like something more basic than a card or
> driver issue. In each case, the cards are correctly detected during
> system boot. For example, the 3Com results in the following output
> during boot (no error messages are printed):
Hello,
Isn't it possible the socket where you put the card is broken or dirty?
I would say the culprit is your machine.
Vilmos
David Cato
> Isn't it possible the socket where you put the card is broken or dirty?
> I would say the culprit is your machine.
That was one of my first thoughts when the problem started. However after
switching cards around and trying different PCI slots, it hasn't made any
difference. The network cards are just flakey no matter which slot I put
them in while the cards that previously occupied the slot were working fine
before and continue to work when installed in the slot in which the network
cards were in previously. Are network cards somehow pickier about their
homes than SCSI or graphics cards?