eth0 and eth1 missing

LYC · 01-25-2014, 11:38 PM

Hi,

We had our server HP DL380 running Redhat version 5 (physical server not VM) relocated and after the server powered up, eth0 and eth1 are missing.

The following was performed with the output:
#ifconfig eth0 up
eth0: unknown interface: no such device

#/etc/init.d/network restart
Bringing up interfac eth0: device eth0 does not seem to present, delaying initialisation

What could be the problem? Is there a way to scan and verify the health of the physical network card and its availability?

Any advise is appreciated.

chantk · 01-25-2014, 11:57 PM

Check your /etc/udev/rules.d/70-persistent-net.rules. The chances are eth0 and eth1 have got mapped to the old NICs.

If you want your new NICs to map to eth0 and eth1, just remove the lines in the file and reboot.

LYC · 01-26-2014, 12:17 AM

Hi,

Don't have 70-persistent-net.rules in /etc/udev/rules.d.

There is no new NIC added.

chantk · 01-26-2014, 12:42 AM

Run "ethtool ethX" and see what you've got (where X is 0 or 1 etc).

No udev? Which version of RedHat are you using? What does "cat /etc/redhat-release" say?

LYC · 01-26-2014, 01:04 AM

Hi,

ethtool eth0
Settings for eth0:
Cannot get device settings: No such device
Cannot get wake-on-lan settings: No such device
Cannot get message level:No such device
Cannot get link status: No such device

ethtool eth1
Settings for eth1:
Cannot get device settings: No such device
Cannot get wake-on-lan settings: No such device
Cannot get message level:No such device
Cannot get link status: No such device

cat /etc/redhat-release
Red Hat Enterprise Linux Server release 5.3 (Tikanga)

chantk · 01-26-2014, 01:15 AM

It's also a good idea to post your dmesg and /var/log/messages.

dmesg and /var/log/message tell a great deal of what's going on with the hardware.

LYC · 01-26-2014, 03:05 AM

Hi,

Refer to attachment.

LYC · 01-26-2014, 03:06 AM

Hi,

2nd part of attachment.

LYC · 01-26-2014, 03:07 AM

Hi,

3rd part of attachment.

LYC · 01-26-2014, 03:08 AM

Hi,

4th part of attachment.

LYC · 01-26-2014, 03:11 AM

Hi,

5th part of attachment.

chantk · 01-26-2014, 05:33 AM

That's quite an effort of screendumping. In fact, your life (and others') will be much easier if you do this:

Code:

dmesg > dmesg.txt

and attach that file.

Anyway, back to your problem. Not sure what you mean by the server was relocated. Hardware change?

Obviously eth0 and eth1 are there. If you had hardware change, it might have to do with SELinux. For debugging purpose, you may want to disable it if it's enabled. Check your /etc/sysconfig/selinux. Make sure it's disabled.. (/selinux should be empty)

vl23 · 01-26-2014, 06:42 AM

lshw -c net ?
lspci if you don't have it, also you should have your datacenter people take a look at the system, chech your iLO hw diagnostic options also.

LYC · 01-26-2014, 10:14 AM

Hi,

There is no ntwork connection. The connection is via KVM, hence, file can not x'fer out and has to use screen shot.

No hardware changes. The server is shifted from one rack to another rack.

more /etc/sysconfig/selinux
SELINUXTYPE=targeted

lshw -c net
bash: command not found

Refer to attachment for lspci output

chantk · 01-26-2014, 12:49 PM

Your dmesg shows the presence of eth0 and eth1. Didn't see it renamed (usually by udev).

What does

Code:

ls -l /sys/class/net

show? If you see ethX or the like, it's most likely something outside the server.

Have you tried disabling selinux?