Fedora stuck on boot spinner after grub

Ask Fedora
,
1

Hello everyone!
I am having problems with fedora after the kernel update, in particular, by running the command “sudo dnf upgrade --refresh”, I updated to kernel 6.10.9-200. Immediately after running it the laptop started to lag and Nautilus crashed. I then tried to reboot to test the new kernel, however at startup I ran into the unpleasant event.

Now, going to boot Fedora with the latest kernel it does not start, it accepts the grub choice, processes it, but while the logo with the spinning wheel is displayed, it freezes and I am forced to restart it forcefully from the button. I also tried with other kernels, but the result is always the same.

I specify that, to use Timeshift to make backups, I renamed the root as @ and the home folder as @home, so at each update I have to edit the grub entry so as to be able to boot without problems.

I also tried to eliminate the “rhgb quiet” options from the grub entry, so as to see what the error was that prevented the boot and I obtained the following in the photo (sorry if the photo is blurry and/or not very clear):

3
31920×2557 785 KB

While, with the “rhgb quiet” option active, but pressing Esc while the wheel is spinning, I get this:

4
41920×1442 657 KB

I also tried using the rescue kernel entry, but I only got more errors.

At this point, I tried to use a live usb to reinstall grub and shim, thinking that the problem was related to some corruption of the related files. I ran the following commands:

mount /dev/nvme0n1p6 /mnt/
mount /dev/nvme0n1p5 /mnt/@/boot
mount /dev/nvme0n1p4 /mnt/@/boot/efi

mount -o bind /dev /mnt/@/dev
mount -o bind /sys /mnt/@/sys
mount -o bind /proc /mnt/@/proc
mount -o bind /run /mnt/@/run
mount -o bind /sys/firmware/efi/efivars /mnt/@/sys/firmware/efi/efivars

chroot /mnt/@

mount -a

dnf reinstall shim-* grub2-*

However, the moment I ran the command to reinstall the packages, I got these errors that scared me:

6
61920×1442 212 KB

I tried creating a folder on @home, to see if it was a disk read/write problem or something else and I got this message: “Error creating directory […]: No space left on device”

I cleaned the user cache folder and deleted some backups of an old phone to get back some space, but it doesn’t work either and I don’t know what to do anymore. I checked the disk status on Windows via CrystalDiskInfo and the nvme is in good condition (99%), so I would rule out the possibility of an end-of-life disk.

I leave the specifications of the laptop I have:

  • Lenovo Legion 5 15ACH6H
  • Fedora 40 (6.10. 9-200) / Windows 11 Pro 23H2 (dual boot)
  • Ryzen 7 5800H + RTX 3070 (I changed the standard 16GB total RAM banks with two 2Rx8 banks for a total of 32GB final)

If further information is needed, just ask me and I will add an edit to this post.
(If you want to see more, the reddit post contains more photos: link)

I thank anyone in advance for the help.

2

Given that you are seeing the systemd logging grub and the shim are working as required. I would not expect reinstalling to be required.

It looks like the first issue is that auditd.service cannot start.

Can you boot into single user mode and get to a terminal to investigate auditd’s failure?

1 Like
3

Hi, thanks for the reply. Unfortunately I cannot log into a single user mode, I tried to add “rw init=/bin/bash” by editing the grub entry and I can use the bash, but something is blocked.
However, I can use the live usb, mount the partitions and use the terminal to execute commands.

Oh, btw, I also ran the command “btrfs check”, the result was positive because no errors/faults/damaged sectors were found.

If you can, tell me what to do and what you need to fully understand my problem and I’ll gladly give me the information. (I also thought to clean install Fedora 40, but that idea is frustrating, I want to develop more knowledge about the fact and be able to understand the root of this problem)

4

It seems you need to add the word “single” to the grub boot.

I am not a btrfs expertt and not sure what you should do next.

5

Sorry for the delay in responding.

When adding “single” by editing the grub entry related to the kernel, there are some things that fail, such as the “systemd-random-seed.service” and the “systemd-update-utmp.service”. Also it says that I can’t access the console because the root account is locked. Could there be something else to try instead?

(I remind, in case it should be useful, that the filesystem is accessible via live usb and can be controlled via chroot)

Since I read that some users had similar problems due to out of memory, I also tried to delete the MATLAB folder (about 12GB of memory), however I didn’t notice any changes in the behavior, in fact the remaining 31.8GB on disk didn’t undergo any changes and remained the same despite the deletion of the program.

6

You may need to balance your btrfs file system so it can write more data to disk.

Search for web for how to do that and how check if its needed.

7

It’s still strange behavior.
Normally, deleting files/folders results in an increase in memory, I don’t want CoW to be the source of the problems.

Anyway, I’ll try to search for something, if I find solutions I’ll post them here

8

Use of mass storage space in BTRFS filesystems is not reported correctly by df:

% doas btrfs filesystem usage /
Overall:
    Device size:		 464.27GiB
    Device allocated:		  59.02GiB
    Device unallocated:		 405.25GiB
    Device missing:		     0.00B
    Device slack:		     0.00B
    Used:			  52.77GiB
    Free (estimated):		 409.71GiB	(min: 207.08GiB)
    Free (statfs, df):		 409.71GiB
    Data ratio:			      1.00
    Metadata ratio:		      2.00
    Global reserve:		 111.09MiB	(used: 0.00B)
    Multiple profiles:		        no

Data,single: Size:55.01GiB, Used:50.55GiB (91.90%)
   /dev/nvme0n1p3	  55.01GiB

Metadata,DUP: Size:2.00GiB, Used:1.11GiB (55.46%)
   /dev/nvme0n1p3	   4.00GiB

System,DUP: Size:8.00MiB, Used:16.00KiB (0.20%)
   /dev/nvme0n1p3	  16.00MiB

Unallocated:
   /dev/nvme0n1p3	 405.25GiB
(base) [gnw3]~% df -H /
Filesystem      Size  Used Avail Use% Mounted on
/dev/nvme0n1p3  499G   57G  440G  12% /

PS: Please avoid using “memory” for “mass storage” or “file system” space so this topic is less likely to be found by people searching for help with memory issues. It can be confusing for those whose native language is not English.

9

Oh I see, thanks for the info on df. I also ran your commands and the output is:

image
image525×529 30.4 KB

Thanks also for the advice on “memory” and “mass storage/filesystem” thing, I will take it into account and try to be more explicit in possible future interactions