newest kernel not running

[Glenn Holmer]

On 04/15/2015 06:05 AM, Chris Luck wrote:
> On 13/04/15 15:58, Glenn Holmer wrote:
>> I'm working on a machine (hosted, for a job, not mine) where, after
>> I ran updates, the latest kernel didn't boot. It's Ubuntu 14.04.2
>> LTS, and I'm seeing kernels 3.13.0-36, -37, -43, and -49 in /boot,
>> but "uname -a" shows that -43 is running. Synaptic tells me that
>> linux-generic's installed version is -49. Also, -49 is at the top of
>> the list in grub.cfg. I can't see the machine booting or pause the
>> GRUB menu because it's a hosted VM.
>>
>> Any ideas how I can diagnose this?

> I solved the matter by following the guidance in message #16 here -
> <http://ubuntuforums.org/showthread.php?t=2250849&page=2>

I finally noticed this output during software updates:

The link /vmlinuz is a damaged link
Removing symbolic link vmlinuz
 you may need to re-run your boot loader[grub]

And sure enough, the vmlinuz and initrd.img links in the root directory
were gone. I re-created them (to point to -49) and re-installed
3.13.0-49; now both the vmlinuz and vmlinuz.old symlinks point to -49,
but after a reboot, "uname -a" still shows 3.13.0-43! I don't know that
GRUB even uses those symlinks, but I'm really stumped at this point.
I've never seen anything quite like this before.

If I were sitting in front of the console, I'd write a custom GRUB cfg
that would force that kernel to load, but as this is a hosted VM, I'm
afraid to lock myself out by rendering it unbootable.

-- 
Glenn Holmer (Linux registered user #16682)
"After the vintage season came the aftermath -- and Cenbe."