Stuck in Grub
Hi, I rebased to an older version of Bazzite (40) from latest (41).
After checking what I wanted I went to rollback to the version I had (41), but upon rebooting I'm now stuck in the Grub command line.
I'm told that I have to load the kernel, but I don't know exactly how to.
102 Replies
Managed to find the kernel but when I boot I get that, what am I missing?
Edit: now getting "Kernel panic - not syncing: No working init found. Try passing init= option to kernel."
:/
does bazzite 40 work?
Managed to get past that by doing initrd and the path to the initramfs, now stuck here.
I'll try and see
Same as above @antheas
i dont know how you managed to break it
š¤·āāļø
Just did ujust rollback after this and that's how it started
I have an usb with bazzite, would reinstalling wipe the data I have?
I'd rather fix this though (if its fixable), as I have already wasted this much time to get here lol
emergency mode kinda hints that something is corrupted, could be the filesystem, idk how to fix btrfs or even diagnose that problem
in emergency mode have you checked smartctl to make sure your drives are not reporting errors?
typing smartctl says command not found
oh so your filesystem isn't even getting mounted
you might have to navigate into one of the sysroots /usr/bin to do it
there are usually two sysroots, one version is the current bazzite, the other is the older version
this is uncharted territory for me, so any help I can get is welcome š
what does ls -la print out?
yea usually navigating in emergency mode requires quite a bit of linux knowledge, or at least know where to look for things and navigate using the CLI
ls -la /
you seem to be inside roots home directory
try
/usr/sbin/smartctl --helpI get
No such file or directorykk give me a second
takes me a bit to remember where those roots are
:okay:
cd /sysroot/ostree/deploy/default/deploy
and then type in ls -la
this directory should have two folders, one for each version of bazzite
here's what mine looks like
I dont have ostree here, only home, root and var
we don't care about the .origin files, its the folders that stores programs, what you are booted into is the plain image with nothing but ostree
ah okay what is the output of
lsblk
i think it might be because you don't have something mountedcommand not found
/usr/bin/lsblk
if that's missing, then i would say your install is toast, boot into a fedora live image and check smartctl -A /dev/nvme0n1 that way
what we want to see with smartctl is to make sure that your drives are not experiencing a hardware failure that rendered your system unusablecan only do /usr/bin/
ah man yea im sorry, wish i was more useful, rpm-ostree based systems like these are very different from what im used to with other distros, idk where else to look for the programs, if you can create a fedora live image stick to boot into to troubleshoot further
ik ostree exists, maybe its in a different partition since I was dualbooting, though havent used Windows since then
other than today to get bazzite again in case this wasnt fixable
I will, shouldnt take long
Just to check, is it Fedora Workstation what Im looking for?
I got Media Writer too
yea
you're not going to install, just use the live media
alright, will @ you when I'm in
@DevilFish303 I'm in
I did smartctl -A /dev/sda5 instead as thats the drive I have bazzite in
it is on its last legs :pain:
about this, it was under root instead
@ me whenever, I'll be around :FeelsStrongMan:
ah yeah thats it
@antheas can you take a look? im not familiar with sata drive codes
pre-fail is i would say a bad sign, what kind of sata is this? tlc? qlc?
i would use a tlc based drive for bazzite, as its heavy on the drive (every update is kind of like a whole OS install, though idk if thats been improved)
like a samsung EVO
its a QLC Seagate SSD
If needed: Seagate BarraCuda Q1 ZA480CV10001
i've seen the "Life" of the drive go down but back when I used Windows, CDI always reported it had good health, just lower %
i use a qlc sata myself, but only for games, ive had it for 4 years or so and only has about 20 or so TBW, its rated for 1300 TBW i think, i hardly swap out games, im paranoid about the write limit on these
I've had it for roughly the same time
if you can boot into windows and check the life of the drive
seagate might have better tooling, samsung has magician on windows for instance
I checked on CDI, said Good 1% š
but will boot on Windows and check with the seagate software
This is what CrystalDiskInfo and HDSentinel report
this is what SeaTools says
Should I go back to the live media? @DevilFish303
nah theres nothing else to do there
its a bit weird that seagate doesnt report a problem, but the other one says theres a day of life left
The 3 report the same thing
though Seagate seems a bit more optimistic with a 2% life
š
is it the 480GB version?
yeah
i've read that when life reaches 0%, the drive goes to read only mode
not sure how true is that as i've never experienced it
you are well under the tbw limit, so thats not it, i think the drive was just flawed
no this is accurate
I see
this is assuming you dont suffer from data loss too
nope
this is the only incident i've had
the drive lasted well past the warranty, funnily enough lmao
I checked in the live fedora and everything was there
So, is any of this fixable seeing that the files are there? https://discord.com/channels/1072614816579063828/1308504095648251945/1308590693614223360 https://discord.com/channels/1072614816579063828/1308504095648251945/1308588120295604346
brb
like i mentioned before emergency mode indicates something is corrupted, ive only seen this happen before on xfs based installs, and when power is cut off during boot that it happens. Repairing the filesystem should be something worth trying but not sure how for btrfs or ostree install in general
I see, how would I go about getting the .txt file from emergency mode?
Would that potentially help?
Here's hoping someone knows :/
Does this mean it's "fine"?
If there's something else I can provide or if there's something I can try in the live media, let me know š
Yes. Installing on a hard drive is not supported
They always do dont they
Backup your data and get a new drive?
I can't atm, so I'd rather fix this if possible
Is your drive busted or not
No?
i thought it was a harddrive lol
okwere once again telling people their hard drive is busted
when its fine
:FeelsStrongMan:
So, what do I do? The files are there, the error I get when booting is "Failed to switch root: /path sysroot doesn't seem to be an OS tree. os-release file is missing." https://discord.com/channels/1072614816579063828/1308504095648251945/1308590693614223360
I'll be away for a bit, please @ me when someone sees and knows what to do š
youre cooked just format already
:FeelsWeirdManW:
Well, since I'm getting no help I'll have to reinstall, thanks I guess
was not saying that, I pinged you to check, it appeared the diagnostic tools were saying it was broke (1 day of life left) since you recommended smartctl in the past i figure you might know how to read those values better than me
also its an ssd, on sata interface, not a hard disk
yea misread that
if it was a hard drive it would have been dead for sure
ok somewhere there you messed up and unsynced grub with the deployments. Dunno how that happened. So the ostree check fails and it bails
so it unsynced by just rolling back?
still don't understand how it even happened tbh
How did you roll back and forth
I rebased to 40 from 41 and everything worked, then when I wanted to go back, I did
ujust rollback
next time I booted I was stuck@Kyle Gospo what does that do
my PC didn't turn off or anything like that in the process
i think it was ujust rollback or rpm-ostree rollback
There was a change in f41 maybe that messes with it
oof
gotta love my luck if that's the case :pain:
brb
So, how do I "sync" it and fix it?
It's honestly 10 times easier to backup your data and reinstall
Hmm
what can I use to back up my home folder?
or just copying to another drive is enough?
I still rather fix it than having to reinstall, but well, just asking to have options
you might be able to fix it by fixing the kernel args of one of the boot options
but its up there
what would be the path?
/boot/loader/entries/ find an entry
and edit its config
you want to edit the kernel params to point to a correct directory if root= is missing
this is how it looks
I don't have aboot.img or aboot.config in that path though
but the rest is there
I have these folders there instead
you tried all the boot options?
try as in?
boot them
I don't get any
what
I have to manually add the kernel and the initramfs
if you mean that, I did both already
oh come on you cant do that
thats why it doesnt work
its fucking bootupd
go fish an old grub and replace the current grub by booting it up to show the old option
the grub is in your /efi partition
inside the fedora folder?
this is all I see
well, it surely didn't work before either, that was the only way I could even get in emergency mode to begin with.
yea
find an older grub, backup those
and replace it
where?
whats this live iso
fedora workstation
i think the easiest thing for you to do now is to boot the bazzite installer, keep your home directory the same and reflash it
theres no point in messing with it anymore
fedora uses a special grub
something happened to it
you deployed with f40, and f41 grub broke
š
Do these look fine?
I'm in the bazzite installation again, everything is still there but I made another partition with gparted just in case...
Can I somehow reuse my old /home?
Well, figured everything out...
I'm missing a few apps (incl. some flatpaks somehow) but it's back