Server Dies After Kernel Upgrade

Home » CentOS » Server Dies After Kernel Upgrade
CentOS 4 Comments

I am running a server with CentOS release 6.4 (Final) and the kernel version of 2.6.32-279.19.1.el6.x86_64 and everything looks ok, but when I do a yum update on the kernel to update it to a newer version “2.6.32-358.11.1.el6”.

It will not restart after the required reboot. It will start to load until the task bar at the bottom gets to the end than it stops loading I have been patient with it unless it requires more time to start but after 10min it was still just sitting at the task bar.

This is a VM machine so I just revert to a previous snapshot, but what should I be looking for that would cause this, and how should I fix it?

Thanks,

4 thoughts on - Server Dies After Kernel Upgrade

  • I recall this happening to me in the past.

    Hit esc when the loading bar starts to run across the screen and see what it stops on.

    Go into single user mode and see if you can look at the logs.

    Can you select the older kernel when the grub pops up?

    You may have to use chkconfig to disable services from starting after you see whats in the way.

  • As Lubomir pointed out yesterday, the latest kernel is vulnerable to a DoS attack, so you should probably use a different one. See Johnny’s message for details, and an untested kernel that you could use instead:

    http://lists.CentOS.org/pipermail/CentOS/2013-July/135671.html

    You can either try ctrl-d or ESC during the splash screen process (I
    don’t remember exactly which one), or you can edit the grub command line when it runs on boot and remove the “rhgb” and “quiet” options from the kernel options.

    –keith

  • There are always 3 kernels available for boot. Only one gets updated. So first check if other two you have are booting properly. Then you can list all available kernels with “yum list kernel
    –showduplicate” and select one and install it like “yum install kernel-2.6.32-358.6.2.el6.CentOS.plus” and see if this one works. If it does, you can use it.

    A thing to check is if you need to recompile VM drivers for new kernels.

  • I cloned the VM that I was having trouble with and installed the new kernel packages I.E kernel, kernel-headers, kernel-firmware, from rpm packs and I was able to get the system to reboot and work no problem. So I went to the production machine and tied using yum to update these same packages, yum also included “bfa-firmware noarch 3.0.3.1-1.el6”. Everything installs ok and I reboot the system, but it takes forever to start hanging for a longtime on processes like sendmail, sm-client, proftpd. I am now wondering if this “bfa-firmware noarch 3.0.3.1-1.el6” package is the problem.

    Thanks,

    Chris

    —–Original Message—

LEAVE A COMMENT