[CentOS-virt] Missing module grub entry in xen-4.4.3-9 & boot issues

Mon Dec 21 10:59:19 UTC 2015
George Dunlap <dunlapg at umich.edu>

On Sun, Dec 20, 2015 at 11:31 AM, Phill Bandelow <phill at onapp.com> wrote:
> HI,
>
> We've started to see several issues with the Xen releases. Going back to
> basics I've used this guide
> https://wiki.centos.org/HowTos/Xen/Xen4QuickStart
>
> Once the install process is complete the grub.conf looks like this:
>
> default=0
> timeout=5
> splashimage=(hd0,0)/boot/grub/splash.xpm.gz
> hiddenmenu
> title CentOS (3.18.21-17.el6.x86_64)
>         root (hd0,0)
>         kernel /boot/xen.gz dom0_mem=1024M,max:1024M cpuinfo com1=115200,8n1
> console=com1,tty loglvl=all guest_loglvl=all
>         module /boot/vmlinuz-3.18.21-17.el6.x86_64 ro root=/dev/vda1
> rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD SYSFONT=latarcyrheb-sun16
> crashkernel=auto  KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet
> title CentOS (2.6.32-573.12.1.el6.x86_64)
>         root (hd0,0)
>         kernel /boot/vmlinuz-2.6.32-573.12.1.el6.x86_64 ro root=/dev/vda1
> rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD SYSFONT=latarcyrheb-sun16
> crashkernel=auto  KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet
>         initrd /boot/initramfs-2.6.32-573.12.1.el6.x86_64.img
> title CentOS (2.6.32-358.el6.x86_64)
>         root (hd0,0)
>         kernel /boot/vmlinuz-2.6.32-358.el6.x86_64 ro
> root=UUID=71b203ab-36ea-4aa8-9ba3-78e4109f0ca4 rd_NO_LUKS rd_NO_LVM
> LANG=en_US.UTF-8 rd_NO_MD SYSFONT=latarcyrheb-sun16 crashkernel=auto
> KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet
>         initrd /boot/initramfs-2.6.32-358.el6.x86_64.img
>
> As you can see the missing module line. If i manually run grub-bootxen.sh
> nothing changes.

Thanks for the detailed report.

The initrd thing is a bug in the upstream script that makes the grub
config; I haven't seen the behavior, so I haven't been able to track
down what the problem is, but other people have[1].

CentOS 7 uses grub2, which comes with its own grub-generation code
which is more reliable.

[1] marc.info/?i=<563670F4.9030701 at karlos.cz>

> Manually adding the missing module entry:
>
> module /boot/initramfs-3.18.21-17.el6.x86_64.img
>
> Causes the following:
>
> https://www.dropbox.com/s/ley5nj29ubwqogc/Screenshot%202015-12-20%2011.10.11.png?dl=0
>
> Disabling APIC gives the following console output:
>
> https://www.dropbox.com/s/a6zgioxl9xpg20y/Screenshot%202015-12-20%2011.15.29.png?dl=0
>
> Now i know this is inside a KVM virtual machine, however we are starting to
> see this issue on at least 7 standard dedicated servers.

You're starting to see that exact error message, you mean, about the
timer not connected to the IOAPIC?

In any case, it would be more helpful if you could manage to get a
serial console output from one of the physical boxes where you've got
a problem.  A lot of servers have built-in serial consoles you can
attach to over the network; otherwise, you can take a look at this
page:

http://wiki.xen.org/wiki/Xen_Serial_Console

Thanks,
 -George