You are not logged in.
Pages: 1
I've managed to get myself into a bit of a predicament. My Arch system is installed on a software RAID0 volume. My /tmp was full for some reason, and it caused some updates to fail so I decided to just reboot. This turned out to be a bad mistake because now I'm getting a kernel panic message and it won't boot.
I have Timeshift backups so I decided to boot a live USB, selected a backup to recover.... and then, to my horror, realised that my system volume wasn't available because it's not on a physical partition and it's not booted. Is there a way around this, or am I completely stuffed?
Last edited by JimDeadlock (2020-04-15 21:19:51)
Offline
Offline
Your description doesn't give much to go on. Specific error messages? /proc/mdstat, mdadm --examine, (or which flavor of raid were you using if not mdadm?)
Unless the hardware is faulty there's a chance you won't even need to recover any backups. The RAID doesn't care about full filesystem; it should still work. You might have to mount, chroot, free up some space and then fixer-upper the botched update.
Offline
Well my initial schoolboy error (this is why I'm in Newbie Corner) was that I was using an Ubuntu live USB, hence the reason I couldn't see the raid volume... so having booted a proper Arch USB I chrooted and tried 'sudo pacman -Syu' but it reported nothing to do. So then I ran timeshift and recovered yesterdays backup, which seemed to go fine. However, I'm still not able to boot.
I'm getting a black screen with a dozen lines, here are the first few relevant lines:
Failed to execute /init (error -13)
Starting init: /sbin/init exists but couldn't execute it (error -13)
Starting init: /bin/init exists but couldn't execute it (error -13)
Starting init: /bin/sh exists but couldn't execute it (error -13)
Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation/admin-guide/init.rst for guidance.
...[snip]I googled an old thread about this, and someone said he resolved it by clearing out his /tmp - I think this was the original cause of my issue, but I've checked /tmp and it's now empty. My init is symlinked to /usr/lib/systemd/systemd and it's executable.
I use EFISTUB as my bootloader if that matters.
Offline
Today's progress...
I discovered that somehow /home/jim was mounted on the root partition (but not the directories within it). This is very strange since /home is supposed to be mounted on a separate partition according to /etc/fstab. Anyway, there were a couple of large files in /home/jim that were causing the root partition to be 100% full. I removed these files, decreasing / to 83% (of 20G). Unfortunately I'm still getting the "Failed to execute /init" error when trying to boot.
I chrooted in to the (broken) root partition from a live USB and ran 'init' to see what would happen. Here's the result, I don't know what it means:
$ init
Cannot determine cgroup we are running in: no medium found
Failed to allocate manager object: no medium foundI'm on the verge of reinstalling Arch but I'd really like to avoid that if at all possible because I have everything the way I like it. Any suggestions would be much appreciated.
Offline
Pages: 1