Rocky Linux wont Boot

Hi all,
I have been building my Rocky Linux 9.4 working environment happily for some time until today… My Rocky Linux 9.4 operating system boots up into a black screen with the following text

[FAILED] Failed to start HQueue Client.
[FAILED] Failed to start HQueue Server.

Here is a link to a video depicting my problem.

Im trying not to panic. I have no idea this problem could happen. I had a fully operational version of Linux that took me weeks to build. I installed Houdini Launcher and using Launcher installed HQeue. I rebooted my PC and now it’s not booting into Linux properly.

If you have any advice, I would greatly appreciate it.
Steve

Are you sure that all the packages you have installed are from the official Rocky repos?

1 Like

I don’t know anything about Houdini, but try going in either under an earlier kernel or rescue mode, and disabling or removing Houdini. It’s apparently not working.

1 Like

Thanks for the suggestions but it all seems beyond my skill level to fix. I keep breaking Linux. I can’t get a stable run to complete my project. Thanks for trying to help.

I was getting warnings about some path running out of disk space, even though I had plenty of disk space available. Maybe I ran out of disk space which buggered the system

I figured out the issue.

booted into a bootable Linux Mint usb drive and ran the boot repair tool. I created a boot info text file. Boot-Info_20240812_0457.txt - Google Drive

Boot Repair says
The /dev/mapper/rl-root (Rocky Linux 9.4 (Blue Onyx)) partition is nearly full. This can prevent to start it. Please use the file browser that just opened to delete unused files (or transfer them to another disk). Close this window when you have finished.

Im not able to clean the partition to free up data because of permissions. So Boot Repair says:
The /dev/mapper/rl-root (Rocky Linux 9.4 (Blue Onyx)) partition is still full. This can prevent to start it (e.g. you may get a Power Manager error).

and then finally Boot Repair says;
Please enable a repository containing the [grub-efi] packages in the software sources of Rocky Linux 9.4 (Blue Onyx) (/dev/mapper/rl-root). Then try again.

So, how can I fix this issue?

Do you think it would be better to just recover my working files and reinstall Linux?

That doesn’t solve the issue though, this insane problem of important partitions running out of space, how do we fix it?

I think there are few options you can try out before reinstalling Linux.

First and easiest thing you can do is to try to find unnecessary files that could be deleted. Your partition scheme only includes /, /home and swap. This means that applications files and logs that are stored within /usr, /opt and /var are stored within your rl-root partition.
Try using the du utility to locate unnecessary files that you could remove:

root@rocky:~# du --max-depth=1 -hc /{usr,opt,var}

Second of all, you could repartition your system. These are your partitions:

lvscan
  ACTIVE            '/dev/rl/swap' [4.00 GiB] inherit
  ACTIVE            '/dev/rl/home' [<3.57 TiB] inherit
  ACTIVE            '/dev/rl/root' [70.00 GiB] inherit

And it seems you have a lot of free space on your /home partition:

                     Avail Use% Mounted on
/dev/mapper/rl-home   2.9T  19% /mnt/boot-sav/mapper/rl-home
/dev/mapper/rl-root    20K 100% /mnt/boot-sav/mapper/rl-root

The quick fix, but one that probably will not help you in the long run, is to delete the swap partition, use the lvm utility to extend the rl-root partition and than expand the xfs filesystem. And than create a swap file on your /home partition instead.
I cannot tell you the exact commands at the moment, but it would go something like that:

root@rocky:~# lvremove /dev/mapper/rl-swap
root@rocky:~# lvextend --size +4G /dev/mapper/rl-root
root@rocky:~# xfs_growfs /
root@rocky:~# dd if=/dev/zero of=/home/swapfile bs=1M count=4096
root@rocky:~# chmod 600 /home/swapfile
root@rocky:~# mkswap /home/swapfile
root@rocky:~# swapon /home/swapfile
root@rocky:~# echo '/home/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab

and delete the old swap entry from /etc/fstab.

The other, more durable way to do it, is to backup your /home partition, remove it with lvremove and extend your rl-root to 200GB or something and recreate your /home.
Unfortunately xfs filesystem cannot be shrinked, therefore you need to remove it.
While you at it, you can rethink your partition scheme and maybe add a separate partion for /var or /usr.
Especially for you only use 20% of your /home. Maybe you could allocate only 1TB to your /home and leave the rest unallocated. This gives you the flexibility to expand and partition that is getting clogged in the future.

2 Likes

Thanks so much for helping me out, I will try and do what you suggested :slight_smile:

The installers default is to create four filesystems: /, /boot, /boot/efi, and /home
The other defaults are to use LVM and XFS and allocate all of the disk for those filesystems.
One cannot shrink existing XFS volume, so the default is rather static (despite having LVM).

The / has files installed from packages. The applications. These do not use huge space, nor change much after installation. The apparent rationale of the default is that users have data (documents) in their home.

However, the / has also /tmp, /var/tmp, and optional stuff under /var. The stuff being system logs and data for services like Apache, podman, and libvirt. Those things the installer cannot anticipate. The admin should customize the install to suite for the known needs.


One can bind mount. For example, have in /etc/fstab:

/home/www_root /var/www/html none bind 0 0

One naturally has to stop the httpd.service and move content of /var/www/html/ into /home/www_root/ before making that mount. Also restorecon -r /var/www/html after mount to recheck the SELinux contexts.

Why such “trick”? When httpd.service is restarted, it will still see files in /var/www/html/ – no need to change the config of the service.

All of this if the /var/www/html is big.

1 Like

Thanks, I will try and get this fixed. Currently, I am my own admin and a newb when it comes to linux.

So in order to fix this, do I start by booting into my ubuntu bootable usb and then try to mount the device with permissions and then follow the instructions or do I boot into Rocky Linux in a recovery mode?

Both ways should work, if you are able to boot into recovery mode. If you use the live usb to boot into ubuntu, you need to mount the /dev/mapper/rl-root somewhere in order to execute xfs_growfs. You do not need to mount the filesystems for the other commands.

1 Like

and if you use the Rocky installation USB and boot it into troubleshooting mode, then it can do that mount for you.

1 Like

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.