Intel NICs, Cisco, autoneg, and borken-ness

Paul Lussier p.lussier at comcast.net
Mon Mar 13 12:05:01 EST 2006


Hi all,

I've got a system up running Debian and a 2.4.22 kernel. 
The primary NIC is:

  eth1      Link encap:Ethernet  HWaddr 00:07:E9:0C:99:2D  
            inet addr:10.0.0.11  Bcast:10.0.255.255  Mask:255.255.0.0


  # netstat -rn
  Kernel IP routing table
  Destination   Gateway     Genmask         Flags   MSS Window  irtt Iface
  10.0.0.0      0.0.0.0     255.255.0.0     U         0 0          0 eth1
  0.0.0.0       10.0.0.1    0.0.0.0         UG        0 0          0 eth1


2 weeks ago I manually configured a second interface on this system to
be on a second network.  I configured the second interface to be on
10.95.0.11/16:


  eth0      Link encap:Ethernet  HWaddr 00:0C:76:14:02:42  
            inet addr:10.95.0.11  Bcast:10.95.255.255  Mask:255.255.0.0

  # netstat -rn
  Kernel IP routing table
  Destination   Gateway     Genmask         Flags   MSS Window  irtt Iface
  10.95.0.0     0.0.0.0     255.255.0.0     U         0 0          0 eth0
  0.0.0.0       10.95.0.1   0.0.0.0         UG        0 0          0 eth0

Using ethtool, I can see that the NIC is hard-wired to be 100baseT/Full:

  # ethtool eth0
  Settings for eth0:

          Supported ports: [ TP ]
          Supported link modes:   10baseT/Half 10baseT/Full 
                                  100baseT/Half 100baseT/Full 
                                  1000baseT/Full 
          Supports auto-negotiation: Yes
          Speed: 100Mb/s
          Duplex: Full
          Port: Twisted Pair
          PHYAD: 0
          Transceiver: internal
          Auto-negotiation: off
          Supports Wake-on: umbg
          Wake-on: d

This NIC is connected to a Cisco 4507 running IOS[1][2], which has
autoneg turned off and that port set to (so I'm told) 100Mb/Full.
When I bring up the interface, I have link on both the switch and the
server, but I can't ping anywhere.  I can't ping the gateway,
10.95.0.1, nor can I ping the other 6 systems I have configured on
this switch, 10.95.33.[23,25,26,27,28,29].  These other 6 systems can
all ping each other and the gateway. The server can ping no one,
including the gateway.  Since this is a /16, the gateway is the same
for all systems, 10.95.0.1.

2 weeks ago, I'm certain I had this interface configured *and* working
just fine.  Today, nothing I've tried seems to work[3].  I've
ifup/ifdown'ed several times, tried manually configuring the interface
via ifconfig, manually set the interface 100Mb/Full, 1000Mb/Full,
autneg, etc. and changing everything on the switch to correspond with
these settings, and nothing works.  I get a link light on the switch
and server *only* when 100Mb/Full is hardwired on the switch, so it
seems logical that I need to set the NIC to that speed, but that
doesn't work either.

dmesg, syslog, and kern.log all show:

 kernel: e1000: eth0 NIC Link is Up 100 Mbps Full Duplex

yet, still no ping...

Any ideas would be greatly appreciated.

Thanks,

Footnotes:
----------
[1] (as opposed to CatIOS)
[2] Full version info:
       Cisco IOS Software, Catalyst 4000            
       L3 Switch Software (cat4000-I9K91S-M), Version 12.2(25)EWA4,
       RELEASE SOFTWARE (fc1)
[3] Given that this is my core NFS server, rebooting is not an option :(

-- 
Seeya,
Paul



More information about the gnhlug-discuss mailing list