There are upcoming maintenance events which may impact our services. Learn more

NIC Down with "e1000e" errors in logs Print

  • 6

If you are running CentOS 6.4 or 6.5, and having problems with the NIC unable to bring up, check logs:-

# tail -100f /var/log/messages

If you see any of the following 'e1000e' errors or any other errors with e1000e:-
e1000e: eth0 NIC Link is Down
e1000e 0000:03:00.0: eth0: Reset adapter
e1000e 0000:03:00.0: eth0: Reset adapter unexpectedly
e1000e 0000:03:00.0: eth0: Timesync Tx Control register not set as expected
e1000e 0000:03:00.0: eth0: Cannot change link characteristics when SoL/IDER is active.

Then this is indicative of the CentOS 6.x kernel Intel e1000e driver issue, as described here:-
http://bugs.centos.org/view.php?id=6810
http://bugs.centos.org/view.php?id=6814

A quick workaround is as follows:-

1. SSH into affected server (if down, try to SSH via internal interface ie 10.x.x.x from a neighboring server).

2. Add pcie_aspm=off kernel parameters to grub config, e.g.:-
# cd /boot/grub
# vim grub.conf
....
        kernel /vmlinuz-2.6.32-431.5.1.el6.x86_64 ro root=/dev/mapper/vg_server21-LogVol00 nomodeset rd_MD_UUID=c8cb6527:97bad21d:75373cd1:0dd9261b  KEYBOARDTYPE=pc KEYTABLE=us LANG=en_US.UTF-8 rd_LVM_LV=vg_server21/LogVol00 rd_LVM_LV=vg_server21/LogVol01 rd_NO_LUKS rd_MD_UUID=ec1c1858:b6c1ad58:6562ea8b:0e8113e2 crashkernel=auto SYSFONT=latarcyrheb-sun16 rd_NO_DM rhgb quiet pcie_aspm=off
....
- (Add pcie_aspm=off at the end of the kernel parameters)

3. Save and reboot
:wq
# shutdown -r now



Was this answer helpful?

« Back