kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

General support questions
man_beach

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by man_beach » 2020/08/02 13:40:32

And how are we meant to do that if we can't boot?

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by TrevorH » 2020/08/02 13:52:51

By booting rescue media, chrooting to your installed system and updating that way.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

man_beach

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by man_beach » 2020/08/03 16:47:35

As I mentioned earlier, the live disc doesn't seem to have a Troubleshooting Rescue option (you only find these things out when you need them) and the minimal disc doesn't have yum. After spending 3 days struggling trying to sort things out, including leaving my ancient netbook on overnight downloading a huge iso only to find in the morning that the checksum was wrong, I finally gave up and reinstalled. Hope it doesn't happen again.

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by TrevorH » 2020/08/03 16:49:23

The minimal iso is definitely all you need since you only need this to boot so that you can start the network and then chroot into your installed system... which _does_ have yum. Or should.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

Starrbuck
Posts: 3
Joined: 2020/08/01 19:15:43

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by Starrbuck » 2020/08/03 18:15:28

That is my experience as well. You have to do some short manual commands to bring the network up to use yum.

jflanakin
Posts: 2
Joined: 2020/05/11 02:25:08

Re: kernel 3.10.0-1127.18.2 / grub 2-2.02-0.86 break UEFI boot

Post by jflanakin » 2020/08/03 18:47:27

I had this happen on my webserver on Friday, 7/31/2020.
Did my regular update and reboot, then surprise surprise I couldn't ssh back in afterwards. Spent a solid day trying to track this down.

My fix:

Code: Select all

Create CentOS 7 Live USB
Boot to Centos 7 Live USB
Choose "Troubleshooting"
Rescue a CentOS Linux System
1) Continue
chroot /mnt/sysimage
df -h
This is just in case you forgot your drives, or to list drives in case you have /boot and /boot/efi on something that isn't /dev/sda. Replace /dev/sda with your drive.

Code: Select all

> grub2-install /dev/sda
	> /sbin/grub2-install: error: /usr/lib/grub/x86_64-efi/modinfo.sh doesn't exist. Please specify —target or —directory
This error that this gives is a symptom of the specific issue. If you can rebuild it like this, you might have a different issue and should probably double check some things. For me, the update removed the old necessary files but never installed the new ones. You may have to set network configuration again. I had to set everything by hand, even had to make a new /etc/resolv file.

Code: Select all

nmtui
	Set network config options manually (IP, gateway, DNS)
vim /etc/resolv (DNS wasn't being set from nmtui)
	Set DNS servers
I also had an issue where yum wasn't working correctly in Rescue mode - it couldn't be found either. Honestly don't know what fixed it. I tried a couple things afterward to no effect, then tried rescue mode again and it worked just fine the second time. YMMV.

Code: Select all

yum makecache
yum reinstall grub2-efi shim
yum install grub2-efi-modules
Reboot
Select Centos …..13 Kernel (just to be safe, could just boot to current kernel)
	press "e"
	add "selinux=0" to the end of the line that starts with "linuxfl" or "linux16"
	press ctrl+x to boot
log in
yum update
yum reinstall selinux-policy-targeted
touch /.autorelabel
systemctl reboot (reboot to earlier kernel again)
yum reinstall kernel
systemctl reboot
After that I was good. I'm planning on cloning the drive and then reinstalling from scratch just to be safe, but I don't run much off my box and can play around like that. It seems stable again, and a couple reboots haven't shown any issue.

Post Reply