Ubuntu Version:
Not sure but I suspect 25.10 (I didn’t do the install and I don’t remember which version was installed)
Desktop Environment (if applicable):
GNOME
Problem Description:
Things were working fine originally, and then they weren’t. The PC was running and the laptop was running in Linux mode, I believe. I came back and it was as if it had rebooted. Since then, I’ve been stuck on initramfs, and I can’t get any further.
Relevant System Information:
It’s an older Lenovo T4080s. It’s dual-booting with Windows 10 on one partition and Linux on the other. It did boot. I had started working with it. Now I just can’t boot in anymore, and I don’t understand why. Windows still boots.
After trying yet again, I managed to do an fdisk -f on the Linux partition and it wen through. So now I have access to my command line and to Wayland (Ibelieve—it’s been a minute since the last time I was on a working Linux system.
Hi
You have a problem. A big problem. Decades ago when one of my installs loaded to BusyBox I re-installed. You might need to do that. Here is a tutorial:
Now here is the bad news. BusyBox is not Linux. BusyBox is a very small operating system that runs lots of commands that are similar to the commands in Linux. This is because both sets of commands are based on similar commands in Unix.
Here is the really bad news. It is the report that UUID=ada7982a-797d-4d28-bd87-0aef1fb39819 does not exist.
I suspect that is the partition Ubuntu is on. UUID= Universally Unique Identifier. It is how a boot manager and operating system identify drives and partitions.
At a very low level that data is still there in binary form. But the markers that give the location of the data throughout the drive are missing. This is how I manage to understand some of this.
Windows loads. This means that you see the Grub boot menu. Select Advanced Options for Ubuntu. The first two lines reference Linux kernel 6.17.0-14. We know that loads to BusyBox. So, select an earlier kernel. I suspect that it will also load to BusyBox. It will be good to know if it does or doesn’t.
You may wish to open another topic on the subject of installing Ubuntu. Also do the same on the subject of retrieving data from a partition. It will avoid confusion.
The picture you show looks like a Ubuntu 24.04 LTS system.
If you were using 25.10; the version of busybox would be 1:1.37.0-4ubuntu1 but yours is much older. The kernel detail (uname -a) you list, shows 24.04 HWE kernel (6.17; backported from 25.10) which has been updated since initial release (24.04.1; update is what .1 ~shows; 24.04 is the only release detail provided by uname)
I agree with what @graymech has already stated; if it was my system I’d just reboot and boot into live media (something I know was good!) & explore all your actual drives you that should be there looking for the partition with UUID shown in your picture and mentioned by @graymech ; as that will have your problem. I may also login to the BIOS/uEFI of your device & see what it detects in your hardware & do some checks there/
You may find a simple file-system check of that partition (what is missing) will fix your boot (if problems are detected & you elect to fix them), but myself, I tend to treat any unexpected problem as a reason to check my hardware, starting with RAM, then disk health (ie. SMART) etc, UNLESS I know what could have caused it (ie. if laptop it could have been left on running using only battery & was not cleanly shutdown, just dying when battery ran out - ie. that is an explanation for why a disk is left clean which can be fixed by a fsck etc I mentioned; do not fsck is one file-system check command, but your install maybe using a file-system that requires a different command).
Your Microsoft Windows, if on a different drive, OR even if on the same actual drive, will be at minimum on a different part of the disk; so I’d be exploring your system for health issues (inc. disk). For me that’s RAM first, PSU (power supply unit; even good components misbehave if fed bad power), disks health etc… If they were clean I’d possibly open the device up & do a cap check (ie. look at motherboard looking for swollen caps, poor air flow etc) or other causes … but that’s just me when I don’t have a reason for the problem, which you appear not to have.
Your blkid output in your initial post clearly shows the system looking for a partition with a UUID that does not exist and also shows 2 Linux partitions (partition 2 and 5) so one of them is your / (root filesystem) partition so you need to determine which it is. You should also check the first partition which is the EFI and should have an ubuntu directory containing a grub.cfg file. Check that for the correct UUID> You then need to put the correct UUID in your grub.cfg and /etc/fstab files.
You can use a ‘live’ Ubuntu (or other Linux) usb to mount and check and modify these files. A UUID does not generally change without some user intervention or formatting. Rebooting or hibernating or similar activities will not change a UUID which is a primary reason they are used.
I’m currenlty not mounting the Windows filesystem.
I rebooted into Windows and it worked.
Rebooted to Linux and I’m at initramfs again. Sigh
Grub gives me 6.17.0-19 and -14.
I tried both and I get initramfs.
Yesterday, however, I did multiple boots and it worked fine.
I suspect that booting Windows messes something up. My multiple reboots into Linux (after doing the fsck from BusyBox and fixing all the inodes) worked fine. Then one Windows reboot and I’m back where I was.
I will create a new live USB later today. Can I install/fix without losing all the data? This is a brand new nvme. I have an older one with my Windows stuff on it so I can always fall back to that if required.
When I exit from initramfs, I get the ALERT! message in my original picture.
The information in your last post shows that the first partition is your EFI partition and partition 5 is your Ubuntu / (root) filesystem partition. What is partition 2 as it is a Linux partition but does not show in your df output meaning it is not mounted.
It was suggested above that you run boot repair with the Create BootInfo Summary option which will give a link you can post here so that members can look at details and help. If will not make any changes unless you as the user tell it to so don’t do that. This would be your best option now.
Do you see a grub menu on boot? If so, you can try the entry below on boot to test. This can be done if you see a grub menu and hit the ‘c’ key on the keyboard which should give you a grub prompt (grub>). If this doesn’t boot, make note of exactly what messages you see as that means your grub.cfg file has errors.
I got the disk running. Internet is working (I’m posting from live disk) yet boot-info doesn’t work. I installed 25.10
$ sudo boot-info -b
Do you want to delete old logs and backups located in /var/log/boot-info ? [yes/no] yes
modprobe: FATAL: Module efivars not found in directory /lib/modules/6.17.0-5-generic
No OS to fix.
Your df output above (from yesterday) shows partition 5 as your Ubuntu root filesystem partition. If you do have a valid partition 2, it was not mounted when you ran the df command which only shows mounted partitions. You would be the only one who knows what that is.
How are you trying to run boot repair? Did you go to the site referenced above and download it on the Ubuntu usb? Did you click on to select the pastebin url to get a report which would show a link to the output for you to post here?
Do you have 25.10 installed on the hard drive? Generally better to use Ubuntu on a live usb with the same version as installed.
The error regarding efivars you posted is common and the location is /sys/firmware/efi/efivars but boot repair looks for it elsewhere.
If you run either sudo fdisk -l OR sudo parted -l from the live Ubuntu usb, do you see the Ubuntu partitions (2 and/or 5)?
What happened when you ran the configile command I suggested earlier
When booting from a live USB, I can’t see partitions nvme0n1p2 and p5.
p5 is the Linux partition. I try to mount it and mount says "probably corrupted filesystem on /dev/nvme0n1p5
I mounted nvme0n1p1 and found the grub.cfg file. It says that the uuid is the one that has the ALERT in the original picture I posted (ada7982a-797d-4d28-bd87-0aef1fb39819).
Here is my pastebin:
Reading it, it looks like I should be replacing my UUID with
Okay, I’m beginning to believe that the problem I’m having is related to booting Windows.
I ended up accepting all the fsck suggestions to repair, and I rebooted on my Ubuntu partition, and it worked like a charm. I rebooted multiple times, three or four times, and it worked every time, and it was extremely quick.
Then I booted Windows 10, waited for it to finish booting, rebooted, and tried to go to the Linux partition, and it failed. I got stuck in initramfs again.
When I run fsck again, it says the drive was not cleanly unmounted.
After some poking around and asking Perplexity, I identified the culprit: Ext2Fsd 0.68.
I removed it, rebooted and… It woiked! Darn, such an easy fix, too.
Don’t use Ext2Fsd 0.68 or earlier versions with latest Ubuntu or
Debian systems. Ext2Fsd 0.68 cannot process EXT4 with 64-BIT mode
enabled, then it could corrupt your data. Very sorry for this
disaster issue, I’m working on an improvement.
Is this Windows utility supposed to read and/or write to/from ext4 file systems?
In doing so, it caused file system damage in Ubuntu because the files were being accessed during a Windows session?
Part of the the name of the utility Ext2Fsd provided the clue
Fsd = File System Damage
Moderation hat on here -
Please mark your thread as “solved” - and many thanks for the informative resolution - check mark box at he bottom of your post #15.
