The other day I set up a fresh installation of Linux Mint, and after a couple hours my desktop environment broke. After restarting, I ended up in busybox with EXT4 file system corruption and I/O errors. I’ve tried everything I could think of (which isn’t a lot, I’m new to Linux) but even creating a fresh install leads to the same corruption and I/O errors. My SSD is the WD Black SN850x, and a test confirmed that there are 0 bad blocks. I’m not sure what to do next, and any help is appreciated.
File system keeps corrupting (need help)
- Thread starter Jac857
- Start date
-
- Tags
- file system ssd
Welcome to the Forum.
The thing I like about Linux mint is it doesn't break and I've never had corrupt system files.
So it's either a bad ISO or user error...did you verify the ISO ?
I find the best way to install Mint...is to select...Erase Disk and Install Linux Mint option. This installs Mint to the whole Drive.
What are your computer specs...Did you try a different Flash Drive or Ventoy.
I have Mint Cinnamon 22.1 running on a 500GB SSD with zero problems...so I can only guess as to what is causing your issues.
The thing I like about Linux mint is it doesn't break and I've never had corrupt system files.
So it's either a bad ISO or user error...did you verify the ISO ?
I find the best way to install Mint...is to select...Erase Disk and Install Linux Mint option. This installs Mint to the whole Drive.
What are your computer specs...Did you try a different Flash Drive or Ventoy.
I have Mint Cinnamon 22.1 running on a 500GB SSD with zero problems...so I can only guess as to what is causing your issues.
Condobloke
Well-Known Member
G'day Jac857, Welcome to Linux.org
What exactly does "my desktop environment broke" ..mean ?
Describe that in some detail , please. What were you doing/trying to do, at that time?
Hi bob466, thank you for your reply. I did verify the ISO and the line matched. It;s possible that I made a mistake during the installation process, so I will try your suggestion to install Mint to the entire Drive. My PC has an AMD Ryzen 7 5800X3D, an RTX 4070, and 4x 16GB of DDR4. I used Rufus initially, but I will try again using Ventoy. Unfortunately I don't have any spare Flash Drives, but if nothing else works I will buy one. Thanks!
Hi Condobloke, that's a great question. I was watching a YouTube video on Firefox, and I noticed that playback had stopped. l looked down at my Panel, and noticed that some of the icons were missing or in the wrong location and clicking on them would do nothing. Afterwards, Firefox closed and then my desktop wallpaper disappeared. I tried to open terminal with commands, and to restart using Ctrl + Alt + Del, and that didn't work so I powered off and that's when I got stuck in BusyBox. Thanks for taking the time to help!
Last edited:
Input/output errors are often hardware problems, but may also be problems of storage with disk, usb, filesystems or even drivers. To get at the issue, it really involves quite a bit of investigation for which the best approach is running of commands in a terminal. If you are new to linux, then this could be a challenge, but the following is a pretty standard way to approach the issue.
Opening a terminal, I'd run the following commands as root, or using sudo, one at a time so one is not overloaded on screen with output:
Code:
dmesg | grep -i error
dmesg | grep -i i/o
dmesg | grep -i reset
dmesg | grep -i bufferThe dmesg command reports what the kernel sees so the point is to see if it is reporting problems. Alternatively one can just run the command
dmesg as root, but then picking through all the detail in perhaps around 1000 lines can be a bit demanding. The above though is where I'd start.Your computer will be just fine to run Mint...rufus won't work in Linux...you could try Etcher.
I suggest Ventoy...if you're using windoze...download the windoze version of Ventoy..it's easy to install to a Flash Drive...then just drag the ISO in.
Hi osprey, thank you for taking the time to reply. I ran the commands using sudo and have attached the the results below.
Got it. Thanks for letting me know; I will try that now
Thanks for that output.
The one problematical looking log is this:
Code:
nvme0: resetting controller due to persistent internal errorIt's a hardware issue. I'm not familiar with the details of it, but I suggest that you google it and see what comes out of that. I did ask AI (gemini) and it produced the following, but I cannot vouch for it ... it's informational:
Code:
The "nvme0: resetting controller due to persistent internal error" message
indicates a critical NVMe SSD failure, often caused by power management conflicts (APST) between Linux and specific hardware, or overheating. It frequently results in system instability or temporary drive disconnection. Fixes include disabling NVMe power management in BIOS, applying kernel parameters, or upgrading firmware.
Common Solutions and Troubleshooting:
Disable Power Management (Most Common Fix): Add nvme_core.default_ps_max_latency_us=0 to your Linux kernel boot parameters (GRUB) to prevent the drive from entering deep sleep states.
PCIe Power Management Settings: Try adding pcie_aspm=off or pcie_port_pm=off to your GRUB settings to disable PCIe power management.
BIOS/Firmware Updates: Update your motherboard BIOS and check for NVMe SSD firmware updates, as this is often a compatibility issue.
Check Hardware: Check if the NVMe drive is overheating, especially on dual-adapter setups. Check physical seating of the drive.I think it's worth checking the drive with smartctl, such as the following command as root or sudo with your disk device name:
Code:
smartctl -a /dev/nvme0n1lsblk command such as this:
Code:
[~]$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sr0 11:0 1 1024M 0 rom
nvme0n1 259:0 0 465.8G 0 disk <----------------------DISK NAME
├─nvme0n1p1 259:1 0 476M 0 part /boot/efi
├─nvme0n1p2 259:2 0 14.9G 0 part [SWAP]
└─nvme0n1p3 259:3 0 450.4G 0 part /The output of the smartctl command above will show whether the disk has PASSED, and also the temperatures of the disk's sensors so they can be checked among a lot of other detail.
I forgot to mention...always disable secure boot. The other thing with SSDs...in Mint you can optimise the SSD to run more efficiently and last for many years.
I have a 500GB SSD...my main SSD in my Tower that's now 7 years old and runs like new. My 14 year old Laptop also has a 500GB SSD but it's about 4 years old.
I have a 500GB SSD...my main SSD in my Tower that's now 7 years old and runs like new. My 14 year old Laptop also has a 500GB SSD but it's about 4 years old.
Thank you for your suggestions, and for helping me to diagnose the issue. It seems that my SSD has passed, and I have attached the results of the smartctl below. For now I will try to apply some of the fixes for what seems to be a power management issue and see if that gets me anywhere. I really appreciate the help!
Attachments
Last edited:
That's really cool! I'm definitely looking forward to getting started with Linux. It seems like there are a lot of benefits over Windows.
Since reinstalling Linux Mint and adding the AI recommended GRUB parameters, I have been able to get to a desktop without encountering any file system corruption or I/O errors in BusyBox, which is great progress. Unfortunately, the desktop environment is still unstable and has crashed multiple times, followed by my SSD not showing up in BIOS or the Boot Menu.
Good to hear of some progress.
It may be worth considering doing a file check which can correct errors in the filesystem. The
fsck command can be used as root or sudo, but it must be run on filesystems that are unmounted. A safe means of doing this is to run a live disk which normally won't have mounted any disks unless you instruct it. For such file checking here, we do it from live disks because it's safe. From the live disk run the following command to check whether the fsck command can see any errors on the disk's partitions. It can be run for each ext4 partition on which there is data:
Code:
fsck -n /dev/<device-name-of-root-partition>lsblk as shown in post #10.If errors are reported in the output of the above command, then to correct everything non-interactively, run as root or sudo for each partition that is shown to have errors:
Code:
fsck -y /dev/<device-name-of-any-partition>As for the BIOS not seeing the SSD, that is unusual to say the least. The BIOS has to see the disk to hand it over to the bootloader. Perhaps check the manufacturer to see if there's a BIOS update.
Edit: In post #12 in the output of smartctl, there is a statistic of 80 unsafe shutdowns. An unsafe shutdown is like when the power goes off and the computer goes off suddenly with it, and the programs that are running don't have time to gracefully shutdown. Users can cause an unsafe shutdown by similar powering off without closing apps as they were intended to close. That sudden stoppage can affect the filesystem causing corruption, hence
fsck is worth running. Some systems will automatically run fsck after an unsafe shutdown, but I can't say how your system runs.
Last edited:
Did you follow the Mint install instructions ?
https://linuxmint-installation-guide.readthedocs.io/en/latest/
As I said...select Erase Disk and Install Linux Mint...also don't Encrypt the Drive.
I don't ever recall seeing the above in the Mint Install instructions.
Does the SSD show in Disks...
Does the SSD show in Computer...
If yes...the SSD is good...it could be another hardware problem.
https://linuxmint-installation-guide.readthedocs.io/en/latest/
As I said...select Erase Disk and Install Linux Mint...also don't Encrypt the Drive.
I don't ever recall seeing the above in the Mint Install instructions.
Does the SSD show in Disks...
Does the SSD show in Computer...
If yes...the SSD is good...it could be another hardware problem.
CaffeineAddict
Well-Known Member
Overall test passed but a ton of info is omitted from the report, smartctl command you need is:
Bash:
sudo smartctl -Ax /dev/nvme0n1 > ~/smart.txtHere is an article from my bookmarks about how to interpret smart report:
I will make 2 suggestions.
1... already mentioned in here was to do a SMART test on the drive and see if you have any errors at all.
2... Do not use encryption, many times that is the cause of more problems than you can imagine. I see it every week from clients.
1... already mentioned in here was to do a SMART test on the drive and see if you have any errors at all.
2... Do not use encryption, many times that is the cause of more problems than you can imagine. I see it every week from clients.
Thank you! I ran the first file check you recommended in a live session, and it came back clean. Should I still try the second fsck command, or since it came back clean does this check mean the file system is alright? Afterwards I updated outdated BIOS to the newest version, and verified that the update worked. However, when I attempted to boot into Linux Mint, I ended up with a purple screen that said "KERNEL PANIC!" and included the following error: "VFS: Unable to mount root fs on unknown-block(0,0)". Also, I'm guessing most of those unsafe shutdowns happened because with the desktop breaking I kept getting stuck unable to Ctrl + Alt +Del and needing to power off with the button.
Last edited:
Similar threads
