On Wed, 27 Jan 2021 at 05:17, GianPiero Puccioni <gianpiero.puccioni@isc.cnr.it> wrote:

Hi,

yesterday my laptop with F32 didn't boot.
It goes in emergency mode and creates a rdsosreport file

I usually don't do this but this time when I installed F I let the system create
the partitions and I think it's LVM with XFS but I'm not sure of the latter and
I am not familiar with this method.

Is there something to do to try to recover something about this, like the
files from /home as of course the USB stick I used for backups went crazy too
and I could recover only a fraction of it. It doesn't seem that it was the HD
that want all bad as the Win10 partition still works.

Often a spinning disk will have a small region that goes bad, so nearly all the

data can be recovered using ddrescue or similar tools. Older spinning disks

have less precise head positioning and become more sensitive to temperature

extremes, so here in Canada, disk errors sometimes disappear if you let a cold

system warm up for a few hours.

You should try to sort out the backups in case ddrescue fails.

If I run lvm_scan I get this:
Scanning devices sda7 for LVM logical volumes fedora_shure/root fedora_shure/swap
ACTIVE '/dev/fedora_shure/root' [50.00 GiB] inherit
ACTIVE '/dev/fedora_shure/swap' [<3.88 GiB] inherit
inactive '/dev/fedora_shure/home' [<166.95 GiB] inherit

and in /dev/fedora_shure there is only root and swap

I'll attach both the rdsosreport and output of journalctl(as suggested by the
error message)

There are disk errors in the journalctl file (the rdsosreport also has journalctl

output). I would check for a loose cable and use ddrescue to create a

copy of the raw drive on a new drive. One of the messages is:

Jan 26 14:06:22 shure kernel: sd 0:0:0:0: [sda] tag#18 Add. Sense: Unrecovered read error - auto reallocate failed

You have a spinning platter disk. These reserve some unallocated blocks which are used to replace blocks

that have errors. "auto reallocate failed" occurs when the spare blocks have all been used. Sometimes the

"errors" are due to a bad cable and the disk is actually fine but needs low-level formatting to recover the

spare blocks (and low-level formatting may not be available to users these days).

Many users complain that their laptop has slowed down when a disk is doing a lot of "bad" block replacements.

If I hear about the slowdown I try to get them a replacement disk. Soon after the slowdown starts the system

will fail to boot but in most cases booting from a live linux distro will allow the user to recover important files.

I used to work with SGI systems that used XFS and applications that beat the disks to death. I just replaced

failed disks and restored the data from backups. The benefit of XFS was that it could reboot after a crash

without a lengthy fsck process. Some data would be lost, but in our applications we were losing data while

systems were down.

Good luck.

George N. White III