CentOS - May 2018 - KVM won't boot after update to 1804

Hi Everyone,

I have a server that I recently applied all updates to to bring it to
1804. After rebooting the server, it would no longer boot and is
instead dropping into a dracut shell. Here's a snippet of the output
printed to the console (lines will likely wrap):

[  184.163787] dracut-initqueue[256]: Warning: dracut-initqueue timeout -
starting timeout scripts
[  184.672525] dracut-initqueue[256]: Warning: dracut-initqueue timeout -
starting timeout scripts
[  185.183111] dracut-initqueue[256]: Warning: dracut-initqueue timeout -
starting timeout scripts
[  185.183395] dracut-initqueue[256]: Warning: Could not boot.
[  185.281196] dracut-initqueue[256]: Warning: /dev/centos_webman03/root does
not exist
[  185.282204] dracut-initqueue[256]: Warning: /dev/centos_webman03/swap does
not exist
[  185.283205] dracut-initqueue[256]: Warning: /dev/mapper/centos_webman11-root
does not exist
         Starting Dracut Emergency Shell...
Warning: /dev/centos_webman03/root does not exist
Warning: /dev/centos_webman03/swap does not exist
Warning: /dev/mapper/centos_webman11-root does not exist

Generating "/run/initramfs/rdsosreport.txt"


Entering emergency mode. Exit the shell to continue.
Type "journalctl" to view system logs.
You might want to save "/run/initramfs/rdsosreport.txt" to a USB stick
or /boot
after mounting them and attach it to a bug report.


dracut:/#


After some monkeying around with the minimal ISO and rescue mode and
google searches, I finally found the combination of commands I needed
to get the system to find the "missing" devices and finish booting:


dracut:/# lvm vgchange -a y
  6 logical volume(s) in volume group "centos_webman11" now active
dracut:/# ln -s /dev/mapper/centos_webman11-root /dev/root
dracut:/# exit
logout
[  469.036109] dracut-initqueue[256]: Warning: Not all disks have been found.
[  469.037349] dracut-initqueue[256]: Warning: You might want to regenerate your
initramfs.
[  OK  ] Started Plymouth switch root service.
[  OK  ] Started Cleanup udevd DB.
[  OK  ] Reached target Switch Root.
         Starting Switch Root...

Welcome to CentOS Linux 7 (Core)!

[snip]


Re-runining grub2-mkconfig doesn't fix the issue: rebooting the server
still resulted in it dropping into a dracut shell. I tried reinstalling
grub2 and that didn't fix it either.

How do I fix this problem? Does anyone have any idea about what caused
this in the first place?



-- 
Ranbir

On Wed, 2018-05-23 at 10:29 -0400, Ranbir wrote:
> 
> Warning: /dev/centos_webman03/root does not exist
> Warning: /dev/centos_webman03/swap does not exist
> Warning: /dev/mapper/centos_webman11-root does not exist
> 
> dracut:/# lvm vgchange -a y
>   6 logical volume(s) in volume group "centos_webman11" now
active
> dracut:/# ln -s /dev/mapper/centos_webman11-root /dev/root
> dracut:/# exit
> logout
> 
> How do I fix this problem? Does anyone have any idea about what
> caused
> this in the first place?
Ah, crap. I just figured out what the problem is. I had changed the
name of the VG and didn't update the grub2 config to reflect that
change. So, of course, no amount of grub2 updating was going to fix the
problem because the devices that had "webman03" truly didn't exist
(they now had "webman11" in them).

After fixing that mistake, the server booted up just fine.

Sorry for the noise!


-- 
Ranbir