[SOLVED] Kernel pinning no effect

Ray_ · Jun 3, 2024

I got one node which pinning has no effect on. Problem is that this node is affected by the 6.8 crashes.

I already tried the steps outlined in https://forum.proxmox.com/threads/unable-to-pin-kernel.111785/ but even the reinit doesn't make a difference.

Code:

root@proxmox3-ng:~# uname -a
Linux proxmox3-ng 6.8.4-2-pve #1 SMP PREEMPT_DYNAMIC PMX 6.8.4-2 (2024-04-10T17:36Z) x86_64 GNU/Linux


root@proxmox3-ng:~# proxmox-boot-tool kernel list
Manually selected kernels:
None.

Automatically selected kernels:
6.5.13-5-pve
6.8.4-2-pve

Pinned kernel:
6.5.13-5-pve

That's also the node I change the rpool disks based on this: https://forum.proxmox.com/threads/replace-512gb-ssds-with-500gb-ssds.143077/

Ray_ · Jun 4, 2024

Booting manually into the older kernel with grub->advanced works, so it has to be a problem with the proxmox-boot-tool

gfngfn256 · Jun 4, 2024

What does proxmox-boot-tool status show?

Ray_ · Jun 4, 2024

gfngfn256 said:
What does proxmox-boot-tool status show?

Nothing out of the ordinary

Code:

root@proxmox3-ng:~# proxmox-boot-tool status
Re-executing '/usr/sbin/proxmox-boot-tool' in new private mount namespace..
System currently booted with legacy bios
3C18-FADA is configured with: grub (versions: 6.5.13-5-pve, 6.8.4-2-pve)
3CD7-1A85 is configured with: grub (versions: 6.5.13-5-pve, 6.8.4-2-pve)

leesteken · Jun 4, 2024

And a proxmox-boot-tool refresh did not help (after adding the pin)?

Ray_ · Jun 4, 2024

leesteken said:
And a proxmox-boot-tool refresh did not help (after adding the pin)?

Sadly not. Also tried unpinning, rebooting and pinning again. That didn't work either.

gfngfn256 · Jun 4, 2024

Maybe try an update-initramfs as this should cause a refresh as per docs, as I guess this is related to your disk replacement.

Ray_ · Jun 4, 2024

gfngfn256 said:
Maybe try an update-initramfs as this should cause a refresh as per docs, as I guess this is related to your disk replacement.

Code:

update-initramfs -u
proxmox-boot-tool reinit
proxmox-boot-tool refresh

Still booting into the unpinned kernel :/

VictorSTS · Jun 4, 2024

I had a similar problem not long ago [1]. Not sure if it applies to your case: you are using root on ZFS but not sure if you are using UEFI. If your system is EFI and you updated at some point from PVE7 to PVE8, you may be missing package systemd-boot. Then proxmox-boot-tool reinit + proxmox-boot-tool refresh should pin the kernel properly.

From the known issues notes of the PVE7 to PVE8 upgrade [2]:

Node Management

Systems booting via UEFI from a ZFS on root setup should install the systemd-boot package after the upgrade.The systemd-boot was split out from the systemd package for Debian Bookworm based releases. It won't get installed automatically upon upgrade from Proxmox VE 7.4 as it can cause trouble on systems not booting from UEFI with ZFS on root setup by the Proxmox VE installer.Systems which have ZFS on root and boot in UEFI mode will need to manually install it if they need to initialize a new ESP (see the output of proxmox-boot-tool status and the relevant documentation).Note that the system remains bootable even without the package installed (the bootloader that was copied to the ESPs during intialization remains untouched), so you can also install it after the upgrade was finished.It is not recommended installing systemd-boot on systems which don't need it, as it would replace grub as bootloader in its postinst script.

[1] https://forum.proxmox.com/threads/proxmox-7-2-efi-zfs-default-kernel-boot.110099/post-617227
[2] https://pve.proxmox.com/wiki/Roadmap#Proxmox_VE_8.0

gfngfn256 · Jun 4, 2024

VictorSTS said:
but not sure if you are using UEFI

His proxmox-boot-tool status output above shows System currently booted with legacy bios

_gabriel · Jun 4, 2024

i bet on system boot from disk/part uuid different from proxmox-boot-tool status show.

edit:
post output of
ls -lha /dev/disk/by-uuid

Ray_ · Jun 4, 2024

_gabriel said:
i bet on system boot from disk/part uuid different from proxmox-boot-tool status show.

edit:
post output of
ls -lha /dev/disk/by-uuid

They seem correct to me

Code:

root@proxmox3-ng:~# ls -lha /dev/disk/by-uuid
lrwxrwxrwx 1 root root  10 Jun  4 15:52 3C18-FADA -> ../../sda2
lrwxrwxrwx 1 root root  10 Jun  4 15:52 3CD7-1A85 -> ../../sdb2

root@proxmox3-ng:~# proxmox-boot-tool status
3C18-FADA is configured with: grub (versions: 6.5.13-5-pve, 6.8.4-2-pve)
3CD7-1A85 is configured with: grub (versions: 6.5.13-5-pve, 6.8.4-2-pve)

root@proxmox3-ng:~# cat /etc/kernel/proxmox-boot-uuids
3C18-FADA
3CD7-1A85

gfngfn256 · Jun 4, 2024

Lets try this step-by-step:

Try the following (and post here ALL the output you get):

proxmox-boot-tool kernel pin 6.5.13-5-pve

Do NOT reboot yet.

proxmox-boot-tool refresh

proxmox-boot-tool kernel list

Reboot

gfngfn256 · Jun 4, 2024

One other thing that maybe worth checking the output of:

efibootmgr -v

_gabriel · Jun 5, 2024

FYI, proxmox-boot-tool edit [vfat-512M-boot-part]\grub\grub.cfg file , like this example :

Code:

diff grub.cfg grub.cfg.pinned
27c27
<    set default="0"
---
>    set default="gnulinux-advanced-d813a054-29dc-49e0-8e28-4f4655b55b08>gnulinux-6.5.13-5-pve-advanced-d813a054-29dc-49e0-8e28-4f4655b55b08"

mount each boot part then check grub\grub.cfg content.

Ray_ · Jun 6, 2024

gfngfn256 said:
Lets try this step-by-step:

Try the following (and post here ALL the output you get):

proxmox-boot-tool kernel pin 6.5.13-5-pve

Do NOT reboot yet.

proxmox-boot-tool refresh

proxmox-boot-tool kernel list

Reboot

Did that, boots into the new kernel

gfngfn256 said:
One other thing that maybe worth checking the output of:

efibootmgr -v

still not using efi

_gabriel said:
FYI, proxmox-boot-tool edit [vfat-512M-boot-part]\grub\grub.cfg file , like this example :

Code:

diff grub.cfg grub.cfg.pinned 27c27 < set default="0" --- > set default="gnulinux-advanced-d813a054-29dc-49e0-8e28-4f4655b55b08>gnulinux-6.5.13-5-pve-advanced-d813a054-29dc-49e0-8e28-4f4655b55b08"

mount each boot part then check grub\grub.cfg content.

Code:

root@proxmox3-ng:~# cat /mnt/boot/grub/grub.cfg | grep default
# from /etc/grub.d and settings from /etc/default/grub
   set default="${next_entry}"
   set default="gnulinux-advanced-0c5a246d8caf017b>gnulinux-6.5.13-5-pve-advanced-0c5a246d8caf017b"

_gabriel · Jun 6, 2024

Ray_ said:

Code:

root@proxmox3-ng:~# cat /mnt/boot/grub/grub.cfg | grep default
# from /etc/grub.d and settings from /etc/default/grub
   set default="${next_entry}"
   set default="gnulinux-advanced-0c5a246d8caf017b>gnulinux-6.5.13-5-pve-advanced-0c5a246d8caf017b"

same result for the two boot partitions ?

Ray_ · Jun 6, 2024

_gabriel said:
same result for the two boot partitions ?

Yeah, both boot drives have the same default value set

gfngfn256 · Jun 6, 2024

Try disconnecting (temporarily) ALL other drives from the host (incl. HDDs & SDDs, USB sticks, SD cards etc.) & boot with ONLY the 2 above drives?

Ray_ · Jun 6, 2024

gfngfn256 said:
Try disconnecting (temporarily) ALL other drives from the host (incl. HDDs & SDDs, USB sticks, SD cards etc.) & boot with ONLY the 2 above drives?

Since this host is a cephnode, I can't do that

[SOLVED] Kernel pinning no effect

Member

Member

Renowned Member

Member

Distinguished Member

Member

Renowned Member

Member

Renowned Member

Node Management​

Renowned Member

Renowned Member

Member

Renowned Member

Renowned Member

Renowned Member

Member

Renowned Member

Member

Renowned Member

Member

Node Management