1

I have an 8-year-old dual boot laptop, but it won't start Ubuntu or Windows. When selecting Ubuntu, I get a BusyBox, that mostly says:

failed command: READ FPDMA QUEUED
blk_update_request: I/O error, dev sda, sector 552058008

enter image description here

Are there any software solutions or is this machine doomed?

I tried for example:

dd if=/dev/sda of=/dev/sda

but it didn't have any effect.

I don't care if I lose all my data, I just want it to be alive again.

I did the smart check via a live CD, and when I first visited the Disks, the /dev/sda was OK, with 57 bad sectors. After running a short and a conveyance, the test failed:

enter image description here

Does that mean that me disk is dead?

All the attributes have an OK assessment in the smart test, so there might be hope. Moreover, if I choose "Check disc for defects" from the live CD, it completes, with no errors found.

gsamaras
  • 615

2 Answers2

3

I did the smart check via a live CD, and when I first visited the Disks, the /dev/sda was OK, with 57 bad sectors. After running a short and a conveyance, the test failed:

In this case the disk is not ok. It's at best dying, and depending on where the sectors are they may effectively make the disk unlikely. The bad sector count will not go down; it will only increase - often relatively fast.

vidarlo
  • 23,497
2

The disk is likely beginning to fail. The SMART test has failed in many ways and the system is reporting IO errors in kernel panics, etc.

Also, this is not a safe command to run:

dd if=/dev/sda of=/dev/sda

That command will read every byte from the disk and write it back to the disk, which is at best useless and at worst destructive.

Another command you can run to likely verify the disk is broken is:

sudo apt install debsums
sudo debsums -s

If anything is printed by the debsums command it means that system files on disk have been corrupted.

Also see this existing question for evaluating the health of your disk:

https://superuser.com/questions/171195/how-to-check-the-health-of-a-hard-drive

Another "easy" way to test is to create a file with random contents, hash it, then copy the file multiple times and observe if any of the copied files have the wrong hash:

dd if=/dev/urandom of=data.bin bs=1G count=1
cp data.bin copy1.bin
cp data.bin copy2.bin
cp data.bin copy3.bin
sha256sum *.bin

If any of the hashes are different the disk has problems storing 4G worth of data. You can increase the sizes to keep testing. Also beware that stressing an already failing disk drastically increases the chances of it having permanent failure.

Kristopher Ives
  • 5,669