A Warning for those running Cent/RHEL 6.x and e1000 NICs

S. Dale Morrey sdalemorrey at gmail.com
Thu Dec 5 16:32:55 MST 2013


Wow, ok I actually spun an AWS RHEL instance that took a dump after an
upgrade and was no longer reachable at all.  I'm wondering if this might be
related.  Fortunately it was a dev box and all the important bits were
backed up.  Still I didn't like being locked out after a dist upgrade like
that.  Not like I can just wander down to the NOC and evaluate it though.


On Thu, Dec 5, 2013 at 2:22 PM, Steve Alligood <steve at betterlinux.com>wrote:

> I have also heard of some people having iscsi initiator issues on boot.
>
>
> On Dec 5, 2013, at 3:08 PM, Ryan Simpkins wrote:
>
> > Long story short:
> > RHEL/CentOS 6.5 doesn't seem fully baked yet. Keep 'er in the oven a bit
> longer.
> >
> > Short story long:
> > If you upgrade to CentOS 6.5 and the 2.6.32-431 kernel, your e1000 based
> Intel
> > NIC will stop working after a few moments upon reboot. You may see the
> > following kernel message:
> >
> > eth0: Timesync Tx Control register not set as expected
> >
> > The work around is to set pcie_aspm=off. I have confirmed this works.
> May this
> > save you a couple of hours of unplanned downtime. :)
> >
> > Bug: http://bugs.centos.org/view.php?id=6810
> >
> > -Ryan
> >
> > /*
> > PLUG: http://plug.org, #utah on irc.freenode.net
> > Unsubscribe: http://plug.org/mailman/options/plug
> > Don't fear the penguin.
> > */
>
>
>
> /*
> PLUG: http://plug.org, #utah on irc.freenode.net
> Unsubscribe: http://plug.org/mailman/options/plug
> Don't fear the penguin.
> */
>


More information about the PLUG mailing list