From Newsgroup: alt.os.linux.mint

On Mon, 11/24/2025 5:37 PM, Felix wrote:

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem? What can I do/use to diagnose/remedy it? thanks

p.s. just another reason Linux is better than Windoze. Win boxes couldn't even see the drive!

sudo apt install smartmontools # Most likely, already installed

sudo smartctl -a /dev/sde # Check drive-reported health

sudo apt install gddrescue # Need a place to put the data (use a spare disk) ... ddrescue

sudo ddrescue -f -n /dev/sde /def/sdf /root/rescue.log # Drive to drive rescue (same sized drive)

sudo xed /root/rescue.log # Examine transfer record, for extent of damage

sudo ddrescue -d -f -r3 /dev/sde /def/sdf /root/rescue.log # Try to recover the remaining damaged sectors

sudo apt install disktype

sudo disktype /dev/sdf # See if the good-quality backup drive, is recognizable.
# Should report the partition setup.

Sometimes, a disk partition, the "envelope" the file system is in
and the file system, are not the same size. It's even possible for
a file system to hang over the end of the drive (which is not good).
Should an OS mount a mis-shaped partition ? IDK. Bad karma.

There is more to disk drives than pretty pictures,
and lots of cool ways it can fail.

I can show you a drive, that has a firmware problem where
the UEFI BIOS issues some sort of command... that causes UEFI
to freeze, with a Seagate 4TB drive. There is some sort of
erroneous response from the drive, that UEFI does not like.
However, if you remove the OS on the 4TB drive, such that
the UEFI "analyze" code is not triggered, the computer starts
fine. This means the drive can only be used as a data drive,
not as an OS drive.

You can use "gnome-disks", to do a read-benchmark of a drive.
There is a menu somewhere in the upper right of gnome-disks,
with the benchmark option. Remember to UNTICK the write-test
option as you do not want gnome-disks to attempt writing while
it benches. The read benchmark is an attempt to see how
sick the drive is (whether it has any "slow-spots" on it).

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Felix <none@not.here> wrote

I noticed a second drive in a Win7 PC was not in File Exploiter, and
didn't appear in file management, so I tried it in a couple of other
PC's, same thing. When I put it in my Linux 22 box it appears unmounted
but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem?

Something has likely died in the electronics card
or the connector isnt connecting properly

What can I do/use to diagnose/remedy it? thanks

Try a new cable. If that doesnt fix it, there is no remedy
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Paul <nospam@needed.invalid> wrote:

On Mon, 11/24/2025 5:37 PM, Felix wrote:

The motor is running since I can feel the vibrations from the
platters spinning. What could be the problem? What can I do/use
to diagnose/remedy it? thanks

p.s. just another reason Linux is better than Windoze. Win boxes
couldn't even see the drive!

sudo apt install smartmontools # Most likely, already installed

sudo smartctl -a /dev/sde # Check drive-reported health

Even if that looks OK, it's more certain to run a self-test with:
sudo smartctl -t long /dev/sde
Then run this every so often to check if it's failed with errors or
succeeded after the estimated completion time:
sudo smartctl -l selftest /dev/sde

Note that the second command may or may not show that the test is
in progress depending on the drive, but it will show up once
complete or failed. Ignore any previous tests still listed there.

Or if the aim is just to get the data off, best to do that first
before deciding whether the drive is faulty or can be reused.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Mon, 11/24/2025 6:39 PM, Felix wrote:

Lawrence DrCOOliveiro wrote:

On Tue, 25 Nov 2025 09:37:11 +1100, Felix wrote:

When I put it in my Linux 22 box it appears unmounted
but LM tries to mount it viz:

https://auslink.info/files/disk2.png

As root:

-a-a-a-a file -s /dev/sde1

peter@ASUS:~$ file -s /dev/sde1
/dev/sde1: no read permission

should report any recognizable filesystem on that partition. If it just
says rCLdatarCY, then yourCOre in trouble ...

-a-a-a-a fdisk -l /dev/sde

peter@ASUS:~$ fdisk -l /dev/sde
fdisk: cannot open /dev/sde: Permission denied

should report some information about the partition setup on the entire
disk.

If the file command reports a valid-looking filesystem, you can also try
manually mounting the disk into a temporary mount point, e.g.

-a-a-a-a mount -o ro /dev/sde1 /mnt

peter@ASUS:~$-a mount -o ro /dev/sde1 /mnt
mount: /mnt: must be superuser to use mount.
-a-a-a-a-a-a dmesg(1) may have more information after failed mount system call.

(note the option to do it reaonly, just in case) and hopefully get a
better message than rCLunknown errorrCY.

does any of that help?

Activities that expect to access the physical layer
of the drive (/dev/zzz), benefit from elevation
such as root (or the ever-present sudo on some distros).

sudo file -s /dev/sde1

sudo fdisk -l /dev/sde

sudo mount -o ro /dev/sde1 /mnt

There is also the disktype program, which is fairly small.

sudo apt install disktype

sudo disktype /dev/sde

and that dumps information about the disk setup, the
partitions on it. Disktype will also process ISO9660 files.

In data recovery, you generally work on the healthy copy you
made of the drive, rather than work on the drive itself. This
also applies to Windows CHKDSK. CHKDSK should not really be
run on a sick disk drive, as the disk may not tolerate the
level of activity involved.

Once a drive is cloned over, you can try the CHKDSK on the clone
if you want. It all depends on whether the signs point to an
actual CHKDSK style problem, or the problem is with the dimensions
of the partitions or other weirdness induced by certain partition
managers and their bad habits.

If nothing seems to be going well with the cloned copy, you can try this.

http://web.archive.org/web/20070101070056/http://www.woundedmoon.org/win32/driverescue19d.html

driverescue19d.zip 1,007,764 bytes
MD5SUM = 63b7e1e8b1701593d5f52c7927d01558

That is for recovering an NTFS partition, from the WinXP era or so.

Normally, modern OSes which maintain hard drives, they do some amount
of background consistency checking. You do not expect large amounts
of "latent faults" to be present in the file system. The drive rescue
program, would be for an NTFS that was never properly maintained.
The drive rescue should have no problem with structural issues, as
a drive exposed to W8/W10/W11 is likely regularly processed by the
OS. Older OSes were processed NTFS, at the direction of the user.

The drive rescue should work best with an NTFS data drive. If you are
working with W11 C: drive, you would be taking quite a chance running it.
The program hails from quite a while ago. The program knows nothing
about reparse points, or the one weird thing I've seen in the $MFT
recently.

But, the drive rescue was free at the time, and it did rescue one disk
for a poster in the WinXP group at the time. There should be lots
of "$39.95 wonder programs" for sale, as well.

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Tue, 25 Nov 2025 10:39:28 +1100, Felix wrote:

Lawrence DrCOOliveiro wrote:

As root:

file -s /dev/sde1

peter@ASUS:~$ file -s /dev/sde1
/dev/sde1: no read permission

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Paul wrote:

On Mon, 11/24/2025 5:37 PM, Felix wrote:

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem? What can I do/use to diagnose/remedy it? thanks

p.s. just another reason Linux is better than Windoze. Win boxes couldn't even see the drive!

sudo apt install smartmontools # Most likely, already installed

sudo smartctl -a /dev/sde # Check drive-reported health

sudo apt install gddrescue # Need a place to put the data (use a spare disk) ... ddrescue

sudo ddrescue -f -n /dev/sde /def/sdf /root/rescue.log # Drive to drive rescue (same sized drive)

sudo xed /root/rescue.log # Examine transfer record, for extent of damage

sudo ddrescue -d -f -r3 /dev/sde /def/sdf /root/rescue.log # Try to recover the remaining damaged sectors

sudo apt install disktype

sudo disktype /dev/sdf # See if the good-quality backup drive, is recognizable.
# Should report the partition setup.

Sometimes, a disk partition, the "envelope" the file system is in
and the file system, are not the same size. It's even possible for
a file system to hang over the end of the drive (which is not good).
Should an OS mount a mis-shaped partition ? IDK. Bad karma.

There is more to disk drives than pretty pictures,
and lots of cool ways it can fail.

I can show you a drive, that has a firmware problem where
the UEFI BIOS issues some sort of command... that causes UEFI
to freeze, with a Seagate 4TB drive. There is some sort of
erroneous response from the drive, that UEFI does not like.
However, if you remove the OS on the 4TB drive, such that
the UEFI "analyze" code is not triggered, the computer starts
fine. This means the drive can only be used as a data drive,
not as an OS drive.

You can use "gnome-disks", to do a read-benchmark of a drive.
There is a menu somewhere in the upper right of gnome-disks,
with the benchmark option. Remember to UNTICK the write-test
option as you do not want gnome-disks to attempt writing while
it benches. The read benchmark is an attempt to see how
sick the drive is (whether it has any "slow-spots" on it).

This is doing my head in. Here's the relative test results. I have no
idea what to make of them (but you do)

https://auslink.info/HD/

Paul

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Rod Speed wrote:

Felix <none@not.here> wrote

I noticed a second drive in a Win7 PC was not in File Exploiter, and
didn't appear in file management, so I tried it in a couple of other
PC's, same thing. When I put it in my Linux 22 box it appears
unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the
platters spinning. What could be the problem?

Something has likely died in the electronics card

That was my thought

or the connector isnt connecting properly

What can I do/use toa diagnose/remedy it? thanks

Try a new cable.

The cable is fine

If that doesnt fix it, there is no remedy

:(
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Tue, 11/25/2025 12:54 AM, Felix wrote:

Paul wrote:

On Mon, 11/24/2025 5:37 PM, Felix wrote:

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem? What can I do/use to diagnose/remedy it? thanks

p.s. just another reason Linux is better than Windoze. Win boxes couldn't even see the drive!

sudo apt install smartmontools-a-a-a # Most likely, already installed

-a-a sudo smartctl -a /dev/sde-a-a-a-a-a-a-a-a # Check drive-reported health >>
sudo apt install gddrescue-a-a-a-a-a-a-a # Need a place to put the data (use a spare disk) ... ddrescue

-a-a sudo ddrescue -f -n-a-a-a-a /dev/sde /def/sdf /root/rescue.log-a-a-a # Drive to drive rescue (same sized drive)

-a-a sudo xed /root/rescue.log-a # Examine transfer record, for extent of damage

-a-a sudo ddrescue -d -f -r3 /dev/sde /def/sdf /root/rescue.log-a-a-a # Try to recover the remaining damaged sectors

sudo apt install disktype

-a-a sudo disktype /dev/sdf-a-a-a-a-a-a-a-a-a-a # See if the good-quality backup drive, is recognizable.
-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a # Should report the partition setup.

Sometimes, a disk partition, the "envelope" the file system is in
and the file system, are not the same size. It's even possible for
a file system to hang over the end of the drive (which is not good).
Should an OS mount a mis-shaped partition ? IDK. Bad karma.

There is more to disk drives than pretty pictures,
and lots of cool ways it can fail.

I can show you a drive, that has a firmware problem where
the UEFI BIOS issues some sort of command... that causes UEFI
to freeze, with a Seagate 4TB drive. There is some sort of
erroneous response from the drive, that UEFI does not like.
However, if you remove the OS on the 4TB drive, such that
the UEFI "analyze" code is not triggered, the computer starts
fine. This means the drive can only be used as a data drive,
not as an OS drive.

You can use "gnome-disks", to do a read-benchmark of a drive.
There is a menu somewhere in the upper right of gnome-disks,
with the benchmark option. Remember to UNTICK the write-test
option as you do not want gnome-disks to attempt writing while
it benches. The read benchmark is an attempt to see how
sick the drive is (whether it has any "slow-spots" on it).

This is doing my head in. Here's the relative test results. I have no idea what to make of them (but you do)

https://auslink.info/HD/

The results don't seem credible.
They're weird looking.

I don't see a power-on hours field, or maybe
I do, and it is lifetime *2 hours* ? Bullshit.

The only thing I can conclude so far, is maybe
the SMART subsystem has reset itself and some
data loss in SMART has occurred or something.

You have a count of "1" in Current Pending Sector count.
Which is suspicious, and those start to show up
near end of life. These seem to happen when the
spares are getting low, and the drive is about
to start reporting CRC errors because there are
no spares to fix that.

The "technical description" of how some of those fields
work, and how they work in practice, is flawed. This has
forced me to describe behaviors in terms of "what I've seen"
versus "what they said in the fine manual".

In Windows, even if a drive won't mount, if the drive is
detect-able, you can do a bad block scan in HDTune. This is
a relatively old program, but should work fine with your
1TB drive. The bad block scan will report sectors with CRC
errors, and the CRC errors tell you how hard it is going to be,
to copy the data off the drive onto a known-good drive. You
won't need to use any of the other tabs. Don't even bother
with the benchmark window, as one CRC error will be enough
to stop a benchmark run.

https://www.hdtune.com/files/hdtune_255.exe

In Linux, this is a clone attempt, where the clone copy
is thrown away. And this is the equivalent of a bad block
scan. The only problem with scanning this way, is the
way that the /root/rescue.log file is encoded. It's
not a simple format reporting one error at a time. The trace
is relatively easy to read, if there is only one error on
the entire disk. But for an unhealthy drive, it would be
just about impossible to read it by eye. The Windows HDTune
pane is easier on the eyeballs.

sudo ddrescue -f -n /dev/sde /def/null /root/rescue.log

It's just possible this disk is unrecoverable. If
the SMART data got torched, there might not be much
left of the poor thing. I've *never* seen nor heard of
SMART getting torched. It always had data in it.
The numbers could be very very large, but it always
had data. How does a drive show it's been powered for
*2* hours. It can't be new, because you have data on it,
and it would take a significant number of power-on hours
just to populate it. And a "new" drive would not be
showing zero reallocated and one Current Pending. That's
not a normal pattern.

A half a dozen more power cycles, that drive might
no longer respond. If you want to get the data off it,
stop toggling the power and concentrate on doing
the ddrescue. That's if ddrescue will continue to run
while hitting all those errors. The disk appears to be
errored somewhere close to the beginning of the disk.

It is your call, on whether this is merely a novelty
observation experiment, or, you are serious about
getting the data off. If I was coming to your house
right now to help, I would be bringing two hard
drives, a known-working 1TB and a known-working larger
one (in case file-at-a-time recovery is attempted).
But the project isn't going to get very far, if the
thing is a mass of errors. Just the time it would take
to reach the other end of the drive, may exceed the number
of hours left before it dies.

Paul


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Tue, 11/25/2025 12:55 AM, Felix wrote:

Rod Speed wrote:

Felix <none@not.here> wrote

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem?

Something has likely died in the electronics card

That was my thought

or the connector isnt connecting properly

What can I do/use toa diagnose/remedy it? thanks

Try a new cable.

The cable is fine

The drive has IDed itself, and it has populated a SMART display.
The error counter for the cable, is not incrementing. Some
of this drive is working. It takes a very large percentage
of functionality, to get this far! Electronics wise, spinning
motor and all, it's functional.

A drive won't respond in any way, unless the heads load
and the ATA command set firmware is loaded off the platter
and into the controller card. This drive has done all those
things. *Some* info *is* coming off the platter. But as
to what the hell is going on in terms of logical state,
I can't tell from this distance. You can hear it spinning,
you "report no clicks of Death", so it's not signaling
"I'm not feeling well" in the usual way. It's a pretty
weird mix of symptoms if you ask me.

A lot of drives appearing to be this mixed up, they would
stop telling you they were a WD10EVDS. They would stop
responding entirely. Why is this disk responding ???

If you want the data off it, do something now.

Paul

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On 25/11/2025 09:11, Paul wrote:

You can hear it spinning,

you "report no clicks of Death",

Where did you see that? The OP just reported vibrations from the
platters spinning.

If you want the data off it, do something now.

Get the disk and put it in a small polythene bag. Put it in a freezer
for at least a couple of hours, preferably overnight. As soon as you
want to try it, take it out of the freezer and put it in a SATA USB
caddy (try to leave as much of the bag on as possible to stop
condensation getting to the disk). Plug the USB lead into your computer
with as little delay as possible, and see what your file manager says
about the disk. If any of the disk is now readable, copy those files to
your computer's HD without delay. It will probably last around 10
minutes before any faulty electronics warm up to prevent further access.

You can try repeating to get more data off, but in the end it will fail.
I've used this a couple of times to get data off a disk - even one with
the click of death. I've got a 1TB Seagate which went down a few months
ago with the click of death, but that unfortunately doesn't respond to
the freezer trick. It's worth trying where everything else has failed.
--
Jeff
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Paul <nospam@needed.invalid> wrote:

On Tue, 11/25/2025 12:54 AM, Felix wrote:

This is doing my head in. Here's the relative test results.
I have no idea what to make of them (but you do)

https://auslink.info/HD/

The results don't seem credible.
They're weird looking.

They look fine to me for a healthy drive.

I don't see a power-on hours field, or maybe
I do, and it is lifetime *2 hours* ? Bullshit.

The power-on hours field on some drives wraps around to zero after
so many hours, so that's not always abnormal.

You have a count of "1" in Current Pending Sector count.
Which is suspicious, and those start to show up
near end of life. These seem to happen when the
spares are getting low, and the drive is about
to start reporting CRC errors because there are
no spares to fix that.

You can see how many reallocated sectors it reports already having
and that's zero! No problem.

[snip]

It is your call, on whether this is merely a novelty
observation experiment, or, you are serious about
getting the data off. If I was coming to your house
right now to help, I would be bringing two hard
drives, a known-working 1TB and a known-working larger
one (in case file-at-a-time recovery is attempted).
But the project isn't going to get very far, if the
thing is a mass of errors. Just the time it would take
to reach the other end of the drive, may exceed the number
of hours left before it dies.

Well he hasn't shown more than typical file system corruption that
can result from a sudden power-off or software crash. But then he
hasn't expressed whether he's interested in the data or wants to
re-use the drive. In the former case, he should stop messing around
looking at SMART data and make a disk image. In the latter case run
a self-test as I suggested before and if that passes then reformat
and get on with using it because the SMART data looks as good as
you'd expect from any used drive.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Tue, 11/25/2025 12:54 AM, Felix wrote:

This is doing my head in. Here's the relative test results.
I have no idea what to make of them (but you do)

https://auslink.info/HD/

The results don't seem credible.
They're weird looking.

They look fine to me for a healthy drive.

I don't see a power-on hours field, or maybe
I do, and it is lifetime *2 hours* ? Bullshit.

The power-on hours field on some drives wraps around to zero after
so many hours, so that's not always abnormal.

You have a count of "1" in Current Pending Sector count.
Which is suspicious, and those start to show up
near end of life. These seem to happen when the
spares are getting low, and the drive is about
to start reporting CRC errors because there are
no spares to fix that.

You can see how many reallocated sectors it reports already having
and that's zero! No problem.

[snip]

It is your call, on whether this is merely a novelty
observation experiment, or, you are serious about
getting the data off. If I was coming to your house
right now to help, I would be bringing two hard
drives, a known-working 1TB and a known-working larger
one (in case file-at-a-time recovery is attempted).
But the project isn't going to get very far, if the
thing is a mass of errors. Just the time it would take
to reach the other end of the drive, may exceed the number
of hours left before it dies.

Well he hasn't shown more than typical file system corruption that
can result from a sudden power-off or software crash. But then he
hasn't expressed whether he's interested in the data or wants to
re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

In the former case, he should stop messing around
looking at SMART data

I've been doing what others suggested

and make a disk image.

how can I make an image of a disk that can't be read?

In the latter case run
a self-test as I suggested

doing that now

before and if that passes then reformat
and get on with using it because the SMART data looks as good as
you'd expect from any used drive.

that's encouraging then, thanks
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Paul wrote:

On Tue, 11/25/2025 12:55 AM, Felix wrote:

Rod Speed wrote:

Felix <none@not.here> wrote

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:
https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png
The motor is running since I can feel the vibrations from the platters spinning. What could be the problem?

Something has likely died in the electronics card

That was my thought

or the connector isnt connecting properly

What can I do/use toa diagnose/remedy it? thanks

Try a new cable.

The cable is fine

The drive has IDed itself, and it has populated a SMART display.
The error counter for the cable, is not incrementing. Some
of this drive is working. It takes a very large percentage
of functionality, to get this far! Electronics wise, spinning
motor and all, it's functional.

A drive won't respond in any way, unless the heads load
and the ATA command set firmware is loaded off the platter
and into the controller card. This drive has done all those
things. *Some* info *is* coming off the platter. But as
to what the hell is going on in terms of logical state,
I can't tell from this distance. You can hear it spinning,
you "report no clicks of Death", so it's not signaling
"I'm not feeling well" in the usual way. It's a pretty
weird mix of symptoms if you ask me.

A lot of drives appearing to be this mixed up, they would
stop telling you they were a WD10EVDS. They would stop
responding entirely. Why is this disk responding ???

If you want the data off it, do something now.

I am trying what is being suggested

Paul

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Jeff Layman wrote:

On 25/11/2025 09:11, Paul wrote:

-aYou can hear it spinning,

you "report no clicks of Death",

Where did you see that? The OP just reported vibrations from the
platters spinning.

If you want the data off it, do something now.

Get the disk and put it in a small polythene bag. Put it in a freezer
for at least a couple of hours, preferably overnight. As soon as you
want to try it, take it out of the freezer and put it in a SATA USB
caddy (try to leave as much of the bag on as possible to stop
condensation getting to the disk). Plug the USB lead into your
computer with as little delay as possible, and see what your file
manager says about the disk. If any of the disk is now readable, copy
those files to your computer's HD without delay. It will probably last around 10 minutes before any faulty electronics warm up to prevent
further access.

You can try repeating to get more data off, but in the end it will
fail. I've used this a couple of times to get data off a disk - even
one with the click of death. I've got a 1TB Seagate which went down a
few months ago with the click of death, but that unfortunately doesn't respond to the freezer trick. It's worth trying where everything else
has failed.

yep, I can try that, thanks. as an alternative I'll go to the local electronics shop and get some freeze spray and give that a try first
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Tue, 11/25/2025 5:55 PM, Felix wrote:

Computer Nerd Kev wrote:

and make a disk image.

how can I make an image of a disk that can't be read?

The partition cannot be mounted.

That limits what it is safe to do.

*******

This leaves a sector-by-sector treatment of the entire disk as an option.

And not just any old sector-by-sector treatment.

The tool must be tolerant of CRC errors (because we think
we can already seen a CRC error right at the beginning of the disk drive).
Many backup programs for example, they hate CRC errors, and attempts
to transfer a disk with them, stop practically instantly.

That is what ddrescue (gddrescue package) is for.

It is a copy of disk dump "dd", where CRC errors are handled
without the transfer stopping.

ddrescue can

transfer from one disk to another disk (works best if disks are same size, because
of a GPT secondary partition table handling issue)

transfer from one disk to a .img file (this can serve as the reference copy of the
bad drive, while you do file-by-file recovery)

The ddrescue command keeps track of which sectors had CRC
errors and were not read on the first pass.

Subsequent runs of the command, as long as they use the same
source-dest pair, can add to the buildup of a "recovered" image.

On the subsequent runs, only the CRC-errored sectors are probed
to see if they are readable. Thus, the approximate runtime
is longest for the first run, shorter for the subsequent runs.

You stop attempting to ddrescue, when the command is not able
to resolve the errored sectors after repeated trials.

*******

Pre-conditioning of drive /dev/sdf # This is the empty, known-to-be-good drive

# You want to erase all the sectors, such that any sectors that are non-zero
# later, you know for sure they came from the bad drive, and are not left over # data from somewhere else. This level of care is needed, if running a file-by-file
# data recovery program later.

sudo fdisk /dev/sdf
p # print details of the recovery drive (there is a "size" field)
q # quit the command without making changes

sudo if=/dev/zero of=/dev/sdf bs=221184 # Usually 221184 divides evenly into drive size
# This could take several hours.

*******

With the minty-fresh /dev/sdf, we now begin the attempt to copy the data.

This is a sector by sector copy, that does not care whether partitions
are intact or whatever.

sudo apt install gddrescue # Need a place to put the data (use a spare disk) ... ddrescue

man ddrescue # Read the instructions, verify the drive names are correct.
# Overwriting the wrong drive, will ruin it.
# Generally, a technician only puts the source and dest
# disks in the machine at this time, for safety.

sudo ddrescue -f -n /dev/sde /dev/sdf /root/rescue.log # Drive to drive rescue (same sized drive)
# First pass gets most of the sectors.
# Copies drive "sde" to drive "sdf".

sudo xed /root/rescue.log # Examine transfer record, for extent of damage
# If, in fact, there were NO crc errors, you can stop here! :-)

sudo ddrescue -d -f -r3 /dev/sde /dev/sdf /root/rescue.log # Try to recover the remaining damaged sectors
# Can be run multiple times, but doing this once
# is usually enough.

The purpose of doing this, is to make sure we haven't lost anything
that is potentially recoverable. The data recovered may be
total garbage, but that's for a later stage of recovery to decide.

1TB bad drive ==> 1TB good drive copy ==> Any size of drive for experiments
with the data. Some recovery programs write
their recovered data to a *separate* disk which
is this disk drive.

You try not to corrupt the "good drive copy", so it is
a reference copy. You would NOT run CHKDSK on the middle
drive. You'd copy the middle drive to the right-hand drive
and run CHKDSK on the right-hand drive.

The reason for this care, is the very next time you turn on the
power on the drive on the left, it might stop responding and
then we can't make any more copies!

Paul


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

But then he hasn't expressed whether he's interested in the data
or wants to re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

Best focus on getting the data off first then because if the drive
is dying it might suddenly stop working entirely after being left
on much longer.

In the former case, he should stop messing around
looking at SMART data

I've been doing what others suggested

and make a disk image.

how can I make an image of a disk that can't be read?

It can be read, otherwise you wouldn't see that there's an NTFS
filesystem on it. The filesystem is corrupt so it can't be
_mounted_. That might be due to the drive failing or the OS messed
it up. Either way if you've got another drive big enough to store
it, make an image using "dd" or "ddrescue" as Paul suggested. I
haven't used the latter, but for dd:

sudo dd if=/dev/sde bs=4096 noerror status=progress of=~/broken_disk.img

WARNING:
Messing up that command can overwrite other drives. You might need
to change of=~/broken_disk.img to point somewhere else if there
isn't enough space on the drive where you home directory is kept
(1TB required), but don't use "of=/dev/[something]".

You can then either run recovery programs on the image, or run them
on the drive if you think it's not dying, but either way you have a
backup image that you can restore if it goes wrong. You can also
compress the backup to avoid needing a full 1TB of space if you're
going to run the recovery software on the drive:

sudo dd if=/dev/sde bs=4096 noerror status=progress | gzip -c > ~/broken_disk.img.gz

In the latter case run a self-test as I suggested

doing that now

Again I would've done that after I got the data off the drive if
that mattered. Although I think in theory you can do both at the
same time.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Computer Nerd Kev <not@telling.you.invalid> wrote:

sudo dd if=/dev/sde bs=4096 noerror status=progress of=~/broken_disk.img

^^^^^^^

sudo dd if=/dev/sde bs=4096 noerror status=progress | gzip -c > ~/broken_disk.img.gz

^^^^^^^
Should be "conv=noerror"
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Wed, 26 Nov 2025 09:55:27 +1100, Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Tue, 11/25/2025 12:54 AM, Felix wrote:

This is doing my head in. Here's the relative test results.
I have no idea what to make of them (but you do)

https://auslink.info/HD/

The results don't seem credible.
They're weird looking.

They look fine to me for a healthy drive.

I don't see a power-on hours field, or maybe
I do, and it is lifetime *2 hours* ? Bullshit.

The power-on hours field on some drives wraps around to zero after
so many hours, so that's not always abnormal.

You have a count of "1" in Current Pending Sector count.
Which is suspicious, and those start to show up
near end of life. These seem to happen when the
spares are getting low, and the drive is about
to start reporting CRC errors because there are
no spares to fix that.

You can see how many reallocated sectors it reports already having
and that's zero! No problem.

[snip]

It is your call, on whether this is merely a novelty
observation experiment, or, you are serious about
getting the data off. If I was coming to your house
right now to help, I would be bringing two hard
drives, a known-working 1TB and a known-working larger
one (in case file-at-a-time recovery is attempted).
But the project isn't going to get very far, if the
thing is a mass of errors. Just the time it would take
to reach the other end of the drive, may exceed the number
of hours left before it dies.

Well he hasn't shown more than typical file system corruption that
can result from a sudden power-off or software crash. But then he
hasn't expressed whether he's interested in the data or wants to
re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

In the former case, he should stop messing around
looking at SMART data

I've been doing what others suggested

and make a disk image.

how can I make an image of a disk that can't be read?

It can be read by ddrescue

In the latter case run
a self-test as I suggested

doing that now

before and if that passes then reformat
and get on with using it because the SMART data looks as good as
you'd expect from any used drive.

that's encouraging then, thanks

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Wed, 26 Nov 2025 10:04:23 +1100, Felix <none@not.here> wrote:

Jeff Layman wrote:

On 25/11/2025 09:11, Paul wrote:

You can hear it spinning,

you "report no clicks of Death",

Where did you see that? The OP just reported vibrations from the
platters spinning.

If you want the data off it, do something now.

Get the disk and put it in a small polythene bag. Put it in a freezer
for at least a couple of hours, preferably overnight. As soon as you
want to try it, take it out of the freezer and put it in a SATA USB
caddy (try to leave as much of the bag on as possible to stop
condensation getting to the disk). Plug the USB lead into your computer
with as little delay as possible, and see what your file manager says
about the disk. If any of the disk is now readable, copy those files to
your computer's HD without delay. It will probably last around 10
minutes before any faulty electronics warm up to prevent further access.

You can try repeating to get more data off, but in the end it will
fail. I've used this a couple of times to get data off a disk - even
one with the click of death. I've got a 1TB Seagate which went down a
few months ago with the click of death, but that unfortunately doesn't
respond to the freezer trick. It's worth trying where everything else
has failed.

yep, I can try that, thanks. as an alternative I'll go to the local electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Felix wrote:

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Tue, 11/25/2025 12:54 AM, Felix wrote:

This is doing my head in. Here's the relative test results.
I have no idea what to make of them (but you do)

https://auslink.info/HD/

The results don't seem credible.
They're weird looking.

They look fine to me for a healthy drive.

I don't see a power-on hours field, or maybe
I do, and it is lifetime *2 hours* ? Bullshit.

The power-on hours field on some drives wraps around to zero after
so many hours, so that's not always abnormal.

You have a count of "1" in Current Pending Sector count.
Which is suspicious, and those start to show up
near end of life. These seem to happen when the
spares are getting low, and the drive is about
to start reporting CRC errors because there are
no spares to fix that.

You can see how many reallocated sectors it reports already having
and that's zero! No problem.

[snip]

It is your call, on whether this is merely a novelty
observation experiment, or, you are serious about
getting the data off. If I was coming to your house
right now to help, I would be bringing two hard
drives, a known-working 1TB and a known-working larger
one (in case file-at-a-time recovery is attempted).
But the project isn't going to get very far, if the
thing is a mass of errors. Just the time it would take
to reach the other end of the drive, may exceed the number
of hours left before it dies.

Well he hasn't shown more than typical file system corruption that
can result from a sudden power-off or software crash. But then he
hasn't expressed whether he's interested in the data or wants to
re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

-a In the former case, he should stop messing around
looking at SMART data

I've been doing what others suggested

and make a disk image.

how can I make an image of a disk that can't be read?

-a In the latter case run
a self-test as I suggested

doing that now

It wouldn't run. gave an error message "can't mount the drive"

before and if that passes then reformat
and get on with using it because the SMART data looks as good as
you'd expect from any used drive.

that's encouraging then, thanks

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

But then he hasn't expressed whether he's interested in the data
or wants to re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

Best focus on getting the data off first then because if the drive
is dying it might suddenly stop working entirely after being left
on much longer.

I don't have another drive to copy to atm

In the former case, he should stop messing around
looking at SMART data

I've been doing what others suggested

and make a disk image.

how can I make an image of a disk that can't be read?

It can be read, otherwise you wouldn't see that there's an NTFS
filesystem on it. The filesystem is corrupt so it can't be
_mounted_. That might be due to the drive failing or the OS messed
it up. Either way if you've got another drive big enough to store
it, make an image using "dd" or "ddrescue" as Paul suggested. I
haven't used the latter, but for dd:

sudo dd if=/dev/sde bs=4096 noerror status=progress of=~/broken_disk.img

WARNING:
Messing up that command can overwrite other drives. You might need
to change of=~/broken_disk.img to point somewhere else if there
isn't enough space on the drive where you home directory is kept
(1TB required), but don't use "of=/dev/[something]".

You can then either run recovery programs on the image, or run them
on the drive if you think it's not dying, but either way you have a
backup image that you can restore if it goes wrong. You can also
compress the backup to avoid needing a full 1TB of space if you're
going to run the recovery software on the drive:

sudo dd if=/dev/sde bs=4096 noerror status=progress | gzip -c > ~/broken_disk.img.gz

I'm not comfortable using the terminal. I'll try the freezing method
when I have access to another drive to copy to

In the latter case run a self-test as I suggested

doing that now

Again I would've done that after I got the data off the drive if
that mattered. Although I think in theory you can do both at the
same time.

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Rod Speed wrote:

On Wed, 26 Nov 2025 10:04:23 +1100, Felix <none@not.here> wrote:

Jeff Layman wrote:

On 25/11/2025 09:11, Paul wrote:

aYou can hear it spinning,

you "report no clicks of Death",

Where did you see that? The OP just reported vibrations from the
platters spinning.

If you want the data off it, do something now.

Get the disk and put it in a small polythene bag. Put it in a
freezer for at least a couple of hours, preferably overnight. As
soon as you want to try it, take it out of the freezer and put it in
a SATA USB caddy (try to leave as much of the bag on as possible to
stop condensation getting to the disk). Plug the USB lead into your
computer with as little delay as possible, and see what your file
manager says about the disk. If any of the disk is now readable,
copy those files to your computer's HD without delay. It will
probably last around 10 minutes before any faulty electronics warm
up to prevent further access.

You can try repeating to get more data off, but in the end it will
fail. I've used this a couple of times to get data off a disk - even
one with the click of death. I've got a 1TB Seagate which went down
a few months ago with the click of death, but that unfortunately
doesn't respond to the freezer trick. It's worth trying where
everything else has failed.

yep, I can try that, thanks. as an alternative I'll go to the local
electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it

Ok
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 27 Nov 2025 05:53:32 +1100, Felix <none@not.here> wrote:

yep, I can try that, thanks. as an alternative I'll go to the local
electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it

Ok

Rod Speed googled that

You can use Gmail for back up
They haxe a 1 month free trail on Google AI Pro (2 TB) https://one.google.com/about/plans?g1_landing_page=0
--
Petzl
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Petzl <petzlx@gmail.com> wrote

Felix <none@not.here> wrote

yep, I can try that, thanks. as an alternative I'll go to the local
electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it

Ok

Rod Speed googled that

Just another of your pathetic little drug crazed drunken fantasys
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Felix wrote:

Computer Nerd Kev wrote:

In the latter case run a self-test as I suggested

doing that now

It wouldn't run. gave an error message "can't mount the drive"

Then you didn't "run a self-test as I suggested" because smartctl
doesn't mount the drive! Like I said:

Even if that looks OK, it's more certain to run a self-test with:
sudo smartctl -t long /dev/sde
Then run this every so often to check if it's failed with errors or
succeeded after the estimated completion time:
sudo smartctl -l selftest /dev/sde

Mounting means reading the file system which we already know is
stuffed up. The point is to figure out whether the drive stuffed it
up because it's dying or it got stuffed up by the OS and the drive
is perfectly fine (until you do something like put it in the
freezer anyway).
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

But then he hasn't expressed whether he's interested in the data
or wants to re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

Best focus on getting the data off first then because if the drive
is dying it might suddenly stop working entirely after being left
on much longer.

I don't have another drive to copy to atm

Well if you value the data then leave it alone until you do.

backup image that you can restore if it goes wrong. You can also
compress the backup to avoid needing a full 1TB of space if you're
going to run the recovery software on the drive:

sudo dd if=/dev/sde bs=4096 noerror status=progress | gzip -c > ~/broken_disk.img.gz

I'm not comfortable using the terminal. I'll try the freezing method
when I have access to another drive to copy to

That's silly when there's nothing to tell you that the drive is at
fault and it could just be that the OS stuffed it up. Freezing might
stuff up a perfectly good drive.

Anyway why ask about getting data off before you had a place to put
it? If you _don't_ really care about the data and are willing to
risk losing it, run the drive's self-test (no "mounting" involved)
and if it passes then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.
Though I've had Windows CHKDSK delete lots of stuff from a drive
before so I'd still recommend making a backup image of it first if
the data is more than curiosity value. If Windows won't run CHKDSK
then you could try using fixntfs on Linux first to mark it "dirty",
but since that's a command-line tool I guess you won't want to.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

good quality good quality
fully operational disk fully operational disk

Bad disk -------> Golden copy disk -------> experiment disk
1TB 1TB 1TB

You do your questionable experiments on the right hand disk.

If the right-hand disk suffers information loss or is
less-good than the middle disk, you clone the middle
disk onto the experiment disk again.

If you have recovery software that scans the middle
disk (such as Photorec), it will ask where you want
the fragments put, and you put the fragments onto the
right hand disk.

Notice that, after the safety copy is made to the middle
disk, we have stopped using the left-hand disk.

*******

If you don't have the setup to attend the left hand disk,
then stop using the left hand disk for now. That's about
the best you can do for it.

Freeze mist isn't necessary, because the disk is being detected,
it's operational, the head is free, the head loads, the
critical data comes off the platter, the disk identifies itself.

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Paul <nospam@needed.invalid> wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

Umm, you snipped:

run the drive's self-test (no "mounting" involved) and if it
passes ^^^^^^^^^

^^^^^^

then leave it connected to a Windows PC while it boots up so
it should automatically CHKDSK it during start-up.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

Agreed, but he doesn't want to do that so...
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Wed, 11/26/2025 8:45 PM, Computer Nerd Kev wrote:

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

The OPs disk showed "two power-on hours".

The SMART is not currently reporting properly.

I have a disk I'm cloning right now, where the same
sort of thing has happened. The Reallocated number previously
reported have been reset. While I was doing a bad
block scan, it was throwing errors like crazy, and
as a final impressive stunt, the drive seemed to reset
itself and when it came back up, the bad block
scan "zoomed ahead" indicating all remaining blocks
had CRC errors.

I currently have the disk being cloned via ddrescue,
and the weird part ? It no longer shows CRC errors,
the disk is behaving in a saintly way, and I have
about one more hour to wait before the copy is complete.

Something is happening to these SMART subsystems, and
on purpose, before the ddrescue started, I was careful
to not use any tool to trigger any SMART activity.
And for some reason, that has changed the symptoms.

I will repeat a previous statement - the drives DO NOT
behave like the hopeful description provided by the manufacturer.
Once your disk is sick, man the life boats, because
you don't know what will happen next. Disks did not always
behave this way. I've also worked on broken disks years
ago, and the behavior (dying right in front of you) was
more predictable. You got lots of clicks of death and
so on. The symptoms were consistent right up to the end.

It seems, up to a certain point, the drive continues to
spare out bad blocks. But, you cannot continue to write
to it and expect to drive it right to the failure point
(no reallocations left). Something has been placed in the
code, which causes mostly random behavior after a certain
point. Yesterday, when I did a bad block scan, about half
the blocks were red. Today, when I use ddrescue -v
there are "0 bad sector error rate: 0 B/sec". Why ?
I have no theory to offer at the moment.

Just, be careful, if you have forced a drive to run too long
in a sick state.

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Felix wrote:

Computer Nerd Kev wrote:

In the latter case run a self-test as I suggested

doing that now

It wouldn't run. gave an error message "can't mount the drive"

Then you didn't "run a self-test as I suggested" because smartctl
doesn't mount the drive! Like I said:

Even if that looks OK, it's more certain to run a self-test with:
sudo smartctl -t long /dev/sde
Then run this every so often to check if it's failed with errors or
succeeded after the estimated completion time:
sudo smartctl -l selftest /dev/sde

Mounting means reading the file system which we already know is
stuffed up. The point is to figure out whether the drive stuffed it
up because it's dying or it got stuffed up by the OS and the drive
is perfectly fine (until you do something like put it in the
freezer anyway).

I did run it as you said, viz..

peter@ASUS:~$ sudo smartctl -t long /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
Sending command: "Execute SMART Extended self-test routine immediately
in off-line mode".
Drive command "Execute SMART Extended self-test routine immediately in off-line mode" successful.
Testing has begun.
Please wait 235 minutes for test to complete.
Test will complete after Wed Nov 26 13:43:26 2025 AEDT
Use smartctl -X to abort test.
peter@ASUS:~$

About 2 pm I looked and there was nothing more in the terminal. I closed
it and then noticed an error message in a window saying the drive
couldn't be mounted, and assumed that the test never actually ran.
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Petzl wrote:

On Thu, 27 Nov 2025 05:53:32 +1100, Felix <none@not.here> wrote:

yep, I can try that, thanks. as an alternative I'll go to the local
electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it

Ok

Rod Speed googled that

yeah, I'm not so sure it would cause condensation. it's used regularly
on electronic circuit boards for fault finding.

You can use Gmail for back up

the drive only has files on it. I use a paid email account linked to my domain. I wouldn't use Gmail if my life depended on it.

They haxe a 1 month free trail on Google AI Pro (2 TB) https://one.google.com/about/plans?g1_landing_page=0

they would need to pay me!
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

But then he hasn't expressed whether he's interested in the data
or wants to re-use the drive.

both. if I can access it I'll get the data of, and if the disk is ok
I'll keep it for non critical use

Best focus on getting the data off first then because if the drive
is dying it might suddenly stop working entirely after being left
on much longer.

I don't have another drive to copy to atm

Well if you value the data then leave it alone until you do.

that's the plan

backup image that you can restore if it goes wrong. You can also
compress the backup to avoid needing a full 1TB of space if you're
going to run the recovery software on the drive:

sudo dd if=/dev/sde bs=4096 noerror status=progress | gzip -c > ~/broken_disk.img.gz

I'm not comfortable using the terminal. I'll try the freezing method
when I have access to another drive to copy to

That's silly when there's nothing to tell you that the drive is at
fault and it could just be that the OS stuffed it up. Freezing might
stuff up a perfectly good drive.

Anyway why ask about getting data off before you had a place to put
it?

I've been asking about getting the drive working.

If you _don't_ really care about the data and are willing to
risk losing it, run the drive's self-test (no "mounting" involved)
and if it passes then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.
Though I've had Windows CHKDSK delete lots of stuff from a drive
before so I'd still recommend making a backup image of it first if
the data is more than curiosity value. If Windows won't run CHKDSK
then you could try using fixntfs on Linux first to mark it "dirty",
but since that's a command-line tool I guess you won't want to.

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

Umm, you snipped:

run the drive's self-test (no "mounting" involved) and if it
passes ^^^^^^^^^

^^^^^^

then leave it connected to a Windows PC while it boots up so
it should automatically CHKDSK it during start-up.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

if it's just an OS stuff up, why isn't there some program I can run to
fix it?

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

Agreed, but he doesn't want to do that so...

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Paul wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

good quality good quality
fully operational disk fully operational disk

Bad disk -------> Golden copy disk -------> experiment disk
1TB 1TB 1TB

You do your questionable experiments on the right hand disk.

If the right-hand disk suffers information loss or is
less-good than the middle disk, you clone the middle
disk onto the experiment disk again.

If you have recovery software that scans the middle
disk (such as Photorec), it will ask where you want
the fragments put, and you put the fragments onto the
right hand disk.

Notice that, after the safety copy is made to the middle
disk, we have stopped using the left-hand disk.

*******

If you don't have the setup to attend the left hand disk,
then stop using the left hand disk for now. That's about
the best you can do for it.

Freeze mist isn't necessary, because the disk is being detected,

not in windows 10/11

it's operational, the head is free, the head loads, the
critical data comes off the platter, the disk identifies itself.

Paul

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 27 Nov 2025 19:12:46 +1100, Felix <none@not.here> wrote:

Paul wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

good quality good quality
fully operational disk fully operational disk

Bad disk -------> Golden copy disk -------> experiment disk
1TB 1TB 1TB

You do your questionable experiments on the right hand disk.

If the right-hand disk suffers information loss or is
less-good than the middle disk, you clone the middle
disk onto the experiment disk again.

If you have recovery software that scans the middle
disk (such as Photorec), it will ask where you want
the fragments put, and you put the fragments onto the
right hand disk.

Notice that, after the safety copy is made to the middle
disk, we have stopped using the left-hand disk.

*******

If you don't have the setup to attend the left hand disk,
then stop using the left hand disk for now. That's about
the best you can do for it.

Freeze mist isn't necessary, because the disk is being detected,

not in windows 10/11

You havent said if its visible in disk manager

it's operational, the head is free, the head loads, the
critical data comes off the platter, the disk identifies itself.

Paul

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 27 Nov 2025 19:08:33 +1100, Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

Umm, you snipped:

run the drive's self-test (no "mounting" involved) and if it
passes ^^^^^^^^^

^^^^^^

then leave it connected to a Windows PC while it boots up so
it should automatically CHKDSK it during start-up.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

if it's just an OS stuff up, why isn't there some program I can run to
fix it?

There maywell be but you want to recover the data
so you cant try wiping the drive to see if that makes
it visible

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

Agreed, but he doesn't want to do that so...

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Felix wrote:

Computer Nerd Kev wrote:

In the latter case run a self-test as I suggested

doing that now

It wouldn't run. gave an error message "can't mount the drive"

Then you didn't "run a self-test as I suggested" because smartctl
doesn't mount the drive! Like I said:

Even if that looks OK, it's more certain to run a self-test with:
sudo smartctl -t long /dev/sde
Then run this every so often to check if it's failed with errors or
succeeded after the estimated completion time:
sudo smartctl -l selftest /dev/sde

Mounting means reading the file system which we already know is
stuffed up. The point is to figure out whether the drive stuffed it
up because it's dying or it got stuffed up by the OS and the drive
is perfectly fine (until you do something like put it in the
freezer anyway).

I did run it as you said, viz..

peter@ASUS:~$ sudo smartctl -t long /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
Sending command: "Execute SMART Extended self-test routine immediately
in off-line mode".
Drive command "Execute SMART Extended self-test routine immediately in off-line mode" successful.
Testing has begun.
Please wait 235 minutes for test to complete.
Test will complete after Wed Nov 26 13:43:26 2025 AEDT
Use smartctl -X to abort test.
peter@ASUS:~$

About 2 pm I looked and there was nothing more in the terminal. I closed
it and then noticed an error message in a window saying the drive
couldn't be mounted, and assumed that the test never actually ran.

You didn't read my instructions properly. You needed to run this
second command to retrieve the results of the self-test which
should have been ready after "Wed Nov 26 13:43:26 2025 AEDT" if you
didn't power off the drive before then:
sudo smartctl -l selftest /dev/sde

You can run that command before the finish time too, to see if it's
failed before the end of the test.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 27 Nov 2025 18:54:51 +1100, Felix <none@not.here> wrote:

Petzl wrote:

On Thu, 27 Nov 2025 05:53:32 +1100, Felix <none@not.here> wrote:

yep, I can try that, thanks. as an alternative I'll go to the local
electronics shop and get some freeze spray and give that a try first

Don't use freeze spray first, that will see condensation
on the logic card that can fuck it

Ok

Rod Speed googled that

yeah, I'm not so sure it would cause condensation.

Anything which is cooled significantly
in other than a perfectly dry environment
will see condensation. Try it with something
where condensation doesnt matter like a glass

it's used regularly on electronic circuit boards for fault finding.

But not with hard drives when you can ensure
that you don't get condensation by putting it in
a plastic bag overnight in the freezer and keeping
it in the plastic bsg in a drive docking thing when
you see if that fixes the problem with the drive
temporarily
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

Umm, you snipped:

run the drive's self-test (no "mounting" involved) and if it
passes ^^^^^^^^^

^^^^^^

then leave it connected to a Windows PC while it boots up so
it should automatically CHKDSK it during start-up.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

if it's just an OS stuff up, why isn't there some program I can run to
fix it?

Since it's NTFS (a M$ format) there's no proper filesystem check
tool on Linux, just fixntfs which will try and make Windows run
CHKDSK at start-up (preferably after you've made a backup). CHKDSK
in Windows is the tool to fix it _if_ the drive is OK.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Computer Nerd Kev <not@telling.you.invalid> wrote:

peter@ASUS:~$ sudo smartctl -t long /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) >> Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org >>
=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
Sending command: "Execute SMART Extended self-test routine immediately
in off-line mode".
Drive command "Execute SMART Extended self-test routine immediately in
off-line mode" successful.
Testing has begun.
Please wait 235 minutes for test to complete.
Test will complete after Wed Nov 26 13:43:26 2025 AEDT
Use smartctl -X to abort test.
peter@ASUS:~$

About 2 pm I looked and there was nothing more in the terminal. I closed
it and then noticed an error message in a window saying the drive
couldn't be mounted, and assumed that the test never actually ran.

You didn't read my instructions properly. You needed to run this
second command to retrieve the results of the self-test which
should have been ready after "Wed Nov 26 13:43:26 2025 AEDT" if you
didn't power off the drive before then:
sudo smartctl -l selftest /dev/sde

You can run that command before the finish time too, to see if it's
failed before the end of the test.

The drive also remembers results from old tests after power-off, so
basically run "sudo smartctl -l selftest /dev/sde" (if the drive is
still at /dev/sde) right away to see if the last self-test had a
result.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 11/27/2025 3:02 AM, Felix wrote:

I've been asking about getting the drive working.

I had a failed disk a couple months ago, in fact
it's the same drive model as is currently in the
Test Machine being ddrescued, and it was clean looking
inside. It would be hard to guess (without resorting
to a microscope), exactly why the surface on these
(while attempting to read) is such a mess.

There used to be disks, where when you opened them up,
the filter pack on the circumference of the HDA, instead
of being a bright white, was covered in dark particulate.
And the wonder back then, is how the disk had been able
to function while that stuff was flying around inside.

But the more recent disks, the perpendicular mode recording
(PMR) ones, the surface doesn't seem to do that. The filter
pack on mine was clean.

The thing is, your disk is identifying itself, which is
a sign it is able to load the ATA command interpreter off
the platter. The head is loaded. You have heard the thing
humming as it spins. These are all signs it is working
from a mechanical perspective. It isn't "stuffed" in
the normal way, where it won't talk to you.

But at a guess, it's been doing a lot of sparing out of sectors.

It's in a bit of a bad mood inside.

There is no obvious way to improve its mood, as without
a pool of spare sectors, there are limited things you
can do to it.

[Picture]

https://i.postimg.cc/GhC86fKg/ddrescue-not-going-well.gif

One of the reasons that disk is in a bad mood, is it is a 250GB
drive with a single platter, and it is a CSS disk, it does not
use a landing ramp. Apparently the landing area near the hub is
laser patterned, so the head won't stick to the platter
("stiction"). The head isn't stuck, but the practice of landing
on the platter, isn't a best practice. If they had build the
disk with a landing ramp, it would last for more service hours.
I thought this practice of doing CSS had stopped, basically,
back when landing ramps were invented. But I guess $0.05 for
a piece of plastic, was too much for them. The landing ramps
are plastic.

Anyway, that picture shows you an attempt by me, to recover
data off one of my rubbish disks. I think I can see partitions
on it, but the partitions won't mount, and it won't even
begin a benchmark run (errors out at the beginning). And
the ddrescue is quoting an absurd number of days to finish
the job (and even then, the percentage of recovered material
might not be sufficient to do anything).

Paul

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Rod Speed wrote:

On Thu, 27 Nov 2025 19:12:46 +1100, Felix <none@not.here> wrote:

Paul wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

aaaaaaaaaaaaaaaaaaaaa good qualityaaaaaaaaaaaaa good quality
aaaaaaaaaaaaaaaaaaaaa fully operational diskaaa fully operational disk

aaa Bad disk -------> Golden copy disk -------> experiment disk
aaaaaa 1TBaaaaaaaaaaaaaaaa 1TBaaaaaaaaaaaaaaaaaaaaaa 1TB

You do your questionable experiments on the right hand disk.

If the right-hand disk suffers information loss or is
less-good than the middle disk, you clone the middle
disk onto the experiment disk again.

If you have recovery software that scans the middle
disk (such as Photorec), it will ask where you want
the fragments put, and you put the fragments onto the
right hand disk.

Notice that, after the safety copy is made to the middle
disk, we have stopped using the left-hand disk.

*******

If you don't have the setup to attend the left hand disk,
then stop using the left hand disk for now. That's about
the best you can do for it.

Freeze mist isn't necessary, because the disk is being detected,

not in windows 10/11

You havent said if its visible in disk manager

In the original post I wrote "I noticed a second drive in a Win7 PC was
not in File Exploiter, and didn't appear in file management" but I see
now that I wrote "file management" instead ofa "disk management"

it's operational, the head is free, the head loads, the
critical data comes off the platter, the disk identifies itself.

aaa Paul

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Rod Speed wrote:

On Thu, 27 Nov 2025 19:08:33 +1100, Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Paul <nospam@needed.invalid> wrote:

On Wed, 11/26/2025 5:58 PM, Computer Nerd Kev wrote:

then leave it connected to a Windows PC while it
boots up so it should automatically CHKDSK it during start-up.

No.

Umm, you snipped:

run the drive's self-test (no "mounting" involved) and if it
passes ^^^^^^^^^

aaa ^^^^^^

then leave it connected to a Windows PC while it boots up so
it should automatically CHKDSK it during start-up.

I've already explained at some point, that in-place repair
of a sick disk with CHKDSK is wrong.

The disk's not sick if SMART shows no failures and it's and passed
the built-in self-test IMHO. The OS might have messed up the data
on it, or power was cut during a write.

if it's just an OS stuff up, why isn't there some program I can run
to fix it?

There may well be but you want to recover the data
so you cant try wiping the drive to see if that makes
it visible

yes

You clone the disk over to known-good materials.
That's your golden copy.

That is your only priority right now. Cloning the
bad disk to a good disk. That's the first step.
The idea is to make sure, before absolutely anything
else happens, you have that copy.

Agreed, but he doesn't want to do that so...

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Felix <none@not.here> wrote:

Felix wrote:

Computer Nerd Kev wrote:

In the latter case run a self-test as I suggested

doing that now

It wouldn't run. gave an error message "can't mount the drive"

Then you didn't "run a self-test as I suggested" because smartctl
doesn't mount the drive! Like I said:

Even if that looks OK, it's more certain to run a self-test with:
sudo smartctl -t long /dev/sde
Then run this every so often to check if it's failed with errors or
succeeded after the estimated completion time:
sudo smartctl -l selftest /dev/sde

Mounting means reading the file system which we already know is
stuffed up. The point is to figure out whether the drive stuffed it
up because it's dying or it got stuffed up by the OS and the drive
is perfectly fine (until you do something like put it in the
freezer anyway).

I did run it as you said, viz..

peter@ASUS:~$ sudo smartctl -t long /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) >> Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org >>
=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
Sending command: "Execute SMART Extended self-test routine immediately
in off-line mode".
Drive command "Execute SMART Extended self-test routine immediately in
off-line mode" successful.
Testing has begun.
Please wait 235 minutes for test to complete.
Test will complete after Wed Nov 26 13:43:26 2025 AEDT
Use smartctl -X to abort test.
peter@ASUS:~$

About 2 pm I looked and there was nothing more in the terminal. I closed
it and then noticed an error message in a window saying the drive
couldn't be mounted, and assumed that the test never actually ran.

You didn't read my instructions properly. You needed to run this
second command to retrieve the results of the self-test which
should have been ready after "Wed Nov 26 13:43:26 2025 AEDT" if you
didn't power off the drive before then:
sudo smartctl -l selftest /dev/sde

You can run that command before the finish time too, to see if it's
failed before the end of the test.

Ok, I'll do it again
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Computer Nerd Kev wrote:

In aus.computers Computer Nerd Kev <not@telling.you.invalid> wrote:

peter@ASUS:~$ sudo smartctl -t long /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) >>> Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org >>>
=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
Sending command: "Execute SMART Extended self-test routine immediately
in off-line mode".
Drive command "Execute SMART Extended self-test routine immediately in
off-line mode" successful.
Testing has begun.
Please wait 235 minutes for test to complete.
Test will complete after Wed Nov 26 13:43:26 2025 AEDT
Use smartctl -X to abort test.
peter@ASUS:~$

About 2 pm I looked and there was nothing more in the terminal. I closed >>> it and then noticed an error message in a window saying the drive
couldn't be mounted, and assumed that the test never actually ran.

You didn't read my instructions properly. You needed to run this
second command to retrieve the results of the self-test which
should have been ready after "Wed Nov 26 13:43:26 2025 AEDT" if you
didn't power off the drive before then:
sudo smartctl -l selftest /dev/sde

You can run that command before the finish time too, to see if it's
failed before the end of the test.

The drive also remembers results from old tests after power-off, so
basically run "sudo smartctl -l selftest /dev/sde" (if the drive is
still at /dev/sde) right away to see if the last self-test had a
result.

Ok. got this:

peter@ASUS:~$ sudo smartctl -l selftest /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF READ SMART DATA SECTION ===
SMART Self-test log structure revision number 1
Num-a Test_Description-a-a-a Status-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a Remaining
LifeTime(hours)-a LBA_of_first_error
# 1-a Extended offline-a-a-a Completed: read failure-a-a-a-a-a-a 90% 4-a-a-a-a-a-a-a-a 6238948

what is that telling us?
--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Paul wrote:

On Thu, 11/27/2025 3:02 AM, Felix wrote:

I've been asking about getting the drive working.

I had a failed disk a couple months ago, in fact
it's the same drive model as is currently in the
Test Machine being ddrescued, and it was clean looking
inside. It would be hard to guess (without resorting
to a microscope), exactly why the surface on these
(while attempting to read) is such a mess.

There used to be disks, where when you opened them up,
the filter pack on the circumference of the HDA, instead
of being a bright white, was covered in dark particulate.
And the wonder back then, is how the disk had been able
to function while that stuff was flying around inside.

But the more recent disks, the perpendicular mode recording
(PMR) ones, the surface doesn't seem to do that. The filter
pack on mine was clean.

The thing is, your disk is identifying itself, which is
a sign it is able to load the ATA command interpreter off
the platter. The head is loaded. You have heard the thing
humming as it spins. These are all signs it is working
from a mechanical perspective. It isn't "stuffed" in
the normal way, where it won't talk to you.

which means, as you said in another post, freezing is irrelevant

But at a guess, it's been doing a lot of sparing out of sectors.

It's in a bit of a bad mood inside.

There is no obvious way to improve its mood, as without
a pool of spare sectors, there are limited things you
can do to it.

[Picture]

https://i.postimg.cc/GhC86fKg/ddrescue-not-going-well.gif

One of the reasons that disk is in a bad mood, is it is a 250GB
drive with a single platter, and it is a CSS disk, it does not
use a landing ramp. Apparently the landing area near the hub is
laser patterned, so the head won't stick to the platter
("stiction"). The head isn't stuck, but the practice of landing
on the platter, isn't a best practice. If they had build the
disk with a landing ramp, it would last for more service hours.
I thought this practice of doing CSS had stopped, basically,
back when landing ramps were invented. But I guess $0.05 for
a piece of plastic, was too much for them. The landing ramps
are plastic.

Anyway, that picture shows you an attempt by me, to recover
data off one of my rubbish disks. I think I can see partitions
on it, but the partitions won't mount, and it won't even
begin a benchmark run (errors out at the beginning). And
the ddrescue is quoting an absurd number of days to finish
the job (and even then, the percentage of recovered material
might not be sufficient to do anything).

Paul

--
Linux Mint 22.2

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Computer Nerd Kev <not@telling.you.invalid> wrote:
The drive also remembers results from old tests after power-off, so
basically run "sudo smartctl -l selftest /dev/sde" (if the drive is
still at /dev/sde) right away to see if the last self-test had a
result.

Ok. got this:

peter@ASUS:~$ sudo smartctl -l selftest /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF READ SMART DATA SECTION ===
SMART Self-test log structure revision number 1
Num Test_Description Status Remaining
LifeTime(hours) LBA_of_first_error
# 1 Extended offline Completed: read failure 90% 4 6238948

what is that telling us?

It failed within the first 10% of the test, so the drive is indeed
dying. Therefore you'll definitely need to find something else to
copy the data onto before you go any further with trying to read
the files on it.
--
__ __
#_ < |\| |< _#
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On 27 Nov 2025 21:51:26 +1000, Computer Nerd Kev
<not@telling.you.invalid> wrote:

In aus.computers Felix <none@not.here> wrote:

Computer Nerd Kev wrote:

In aus.computers Computer Nerd Kev <not@telling.you.invalid> wrote:
The drive also remembers results from old tests after power-off, so
basically run "sudo smartctl -l selftest /dev/sde" (if the drive is
still at /dev/sde) right away to see if the last self-test had a
result.

Ok. got this:

peter@ASUS:~$ sudo smartctl -l selftest /dev/sde
[sudo] password for peter:
smartctl 7.4 2023-08-01 r5530 [x86_64-linux-6.8.0-88-generic] (local build) >> Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org >>
=== START OF READ SMART DATA SECTION ===
SMART Self-test log structure revision number 1
Num Test_Description Status Remaining
LifeTime(hours) LBA_of_first_error
# 1 Extended offline Completed: read failure 90% 4 6238948 >>
what is that telling us?

It failed within the first 10% of the test, so the drive is indeed
dying. Therefore you'll definitely need to find something else to
copy the data onto before you go any further with trying to read
the files on it.

Time for euthanasia of your drives data I'm afraid.
The SSD drive might be usable after google how to revive
Otherwise
SSD Drives good ones at a best price are from TEMU but look for a
quality maker "KODAK"
--
Petzl
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 11/27/2025 7:39 PM, Petzl wrote:

On 27 Nov 2025 21:51:26 +1000, Computer Nerd Kev
<not@telling.you.invalid> wrote:

In aus.computers Felix <none@not.here> wrote:

what is that telling us?

It failed within the first 10% of the test, so the drive is indeed
dying. Therefore you'll definitely need to find something else to
copy the data onto before you go any further with trying to read
the files on it.

Time for euthanasia of your drives data I'm afraid.
The SSD drive might be usable after google how to revive

Otherwise SSD Drives good ones at a best price are from TEMU but look for a quality maker "KODAK"

TEMU ??? You've got to be joking.

They wouldn't even pay their delivery drivers back wages here.

And some companies sell their trade name to others,
so a "Duracell" SSD, the staff at the Duracell corporation
would know nothing about the details of the item. The same
would go for "Kodak". Just a name.

https://en.wikipedia.org/wiki/Kodak

"Post-bankruptcy

... Kodak has licensed its brand to several other companies.
"

The suggestion here, is the Kodak SSD is a rebranded Emtec. [April 2019]

https://hardforum.com/threads/hot-and-weird-kodak-branded-480gb-ssd-2-5-39-99-woot-amazon.1980051/

https://web.archive.org/web/20150502181048/https://www.pureoverclock.com/Review-detail/emtec-international-power-plus-x150-240gb-solid-state-drive-review-2/2/

I would recommend something that has more reviews,
and is more mainstream, making it easier to judge
whether you want to go near it with a barge pole.

Modern SSDs use TLC (3-bit) flash or QLC (4-bit) flash.
Both can get mushy after three months. A good firmware
will re-write the blocks in the background, to give
the impression the drive is not mushy. That makes it
look like it was more of an MLC (2-bit) drive, but
at the cost of burning up some wear life.

SLC and MLC (1-bit and 2-bit) flash, make excellent
SSD drives, where write performance is constant from
end to end. This is why they don't make SLC and MLC flash
for consumers any more, preferring instead to sell TLC mush.
But with the right firmware, you can hide some of what is
going on, and this is what we're paying for.

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Thu, 27 Nov 2025 23:06:43 -0500, Paul <nospam@needed.invalid>
wrote:

On Thu, 11/27/2025 7:39 PM, Petzl wrote:

On 27 Nov 2025 21:51:26 +1000, Computer Nerd Kev
<not@telling.you.invalid> wrote:

In aus.computers Felix <none@not.here> wrote:

what is that telling us?

It failed within the first 10% of the test, so the drive is indeed
dying. Therefore you'll definitely need to find something else to
copy the data onto before you go any further with trying to read
the files on it.

Time for euthanasia of your drives data I'm afraid.
The SSD drive might be usable after google how to revive

Otherwise SSD Drives good ones at a best price are from TEMU but look for a >> quality maker "KODAK"

TEMU ??? You've got to be joking.

They wouldn't even pay their delivery drivers back wages here.

And some companies sell their trade name to others,
so a "Duracell" SSD, the staff at the Duracell corporation
would know nothing about the details of the item. The same
would go for "Kodak". Just a name.

https://en.wikipedia.org/wiki/Kodak

"Post-bankruptcy

... Kodak has licensed its brand to several other companies.
"

The suggestion here, is the Kodak SSD is a rebranded Emtec. [April 2019]

https://hardforum.com/threads/hot-and-weird-kodak-branded-480gb-ssd-2-5-39-99-woot-amazon.1980051/

https://web.archive.org/web/20150502181048/https://www.pureoverclock.com/Review-detail/emtec-international-power-plus-x150-240gb-solid-state-drive-review-2/2/

I would recommend something that has more reviews,
and is more mainstream, making it easier to judge
whether you want to go near it with a barge pole.

Modern SSDs use TLC (3-bit) flash or QLC (4-bit) flash.
Both can get mushy after three months. A good firmware
will re-write the blocks in the background, to give
the impression the drive is not mushy. That makes it
look like it was more of an MLC (2-bit) drive, but
at the cost of burning up some wear life.

SLC and MLC (1-bit and 2-bit) flash, make excellent
SSD drives, where write performance is constant from
end to end. This is why they don't make SLC and MLC flash
for consumers any more, preferring instead to sell TLC mush.
But with the right firmware, you can hide some of what is
going on, and this is what we're paying for.

Paul

Fair enough a Koack SSD drive is a lot cheaper, qualty seems OK to me
I have two KODK 1 Tetrabyts SSD USB drives for around two years
TEMU also have mlc ssd PC drives but a dearer price
$299
https://imgbb.com/
--
Petzl
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Mon, 11/24/2025 5:37 PM, Felix wrote:

I noticed a second drive in a Win7 PC was not in File Exploiter, and didn't appear in file management, so I tried it in a couple of other PC's, same thing. When I put it in my Linux 22 box it appears unmounted but LM tries to mount it viz:

https://auslink.info/files/disk1.png
https://auslink.info/files/disk2.png

The motor is running since I can feel the vibrations from the platters spinning. What could be the problem? What can I do/use to diagnose/remedy it? thanks

p.s. just another reason Linux is better than Windoze. Win boxes couldn't even see the drive!

I found another program that claims to be a cloning tool.
It is like ddrescue, except it uses a separate driver for
the "bad" disk, that driver is capable of keeping the
bad driver operating, it truncates error recovery on
the drive and so on. I'm running it right now on
my sample "bad drive" which is really sick, and it
seems to manage 3MB/sec transfer rate off a bad disk.

https://www.hddsuperclone.com/hddlivecd/download

The Pro version of the product, has become free, as of 2022 or so.

The download is hosted on Google Drive, and you have to be
careful to click and get down to the bottom level, to do the
download correctly. If it says it is "zipping something",
you need to carefully click the folder and descend lower to
where the desired ISO is.

Name: xubuntu-18.04.5-desktop-hddlive.2022_07_24.x64.1.5G.iso
Size: 1,495,748,608 bytes (1426 MiB)
SHA1: E8E71E58151DE99BDEE7373C1FD3549561ED3087

That's a legacy boot release of OS, with HDDSuperClone executable
on the desktop of the LiveDVD session.

To use it, when you boot the LiveDVD, there are a number of (GRUB) menu items. You don't use the first one, which is just the normal Xubuntu Live session.
The second entry in the menu, mentions "hiding a drive", and what that
does, is prevents the normal Linux driver from attaching to the sick disk.
(I used the custom "direct AHCI" driver.) This is my hardware setup for the test.

Port 2 600MB/sec 2TB WD Black drive for accepting clone copy

Port 4 300MB/sec 250MB Seagate sick drive, needing to be cloned off <=== source drive

Where it says to "edit the value where the X" is then, I place the digit "4"
in its place. And libata notches out that port and does not attach to it
during the boot.

when the desktop appears, and HDDSuperClone is run from the desktop icon,
it has a monitor terminal which logs activity, and it uses a copy of gcc
to compile a custom driver for port 4. That driver then binds to port 4
and runs the 250MB Seagate.

As a practical person, how can you tell what the port number is ???
If you use the popup boot key on the computer, and the popup boot
menu appears, the port numbers are part of the boot menu display.
In there I can see the 2TB drive with "2" showing and the 250GB
drive with the "4" showing, as well as an entry for the LiveDVD
I've inserted in the tray.

I use the popup boot (F8 on Asus, F11 on MSI) and bring up the popup
boot menu, make note of the port numbers, then I cursor down to the
Optical drive and boot from the DVD. Then, once the GRUB menu appears,
I select the second entry and edit the "X" part of the command and
change that to "4" in my case.

There is a 79 page PDF manual with the product.

Name: hddsuperclone.pdf
Size: 299,830 bytes (292 KiB)
SHA1: 0ECD9F11413381AF8973C7B2271084C1EB059DC9

D:\xubuntu-18.04.5-desktop-hddlive.2022_07_24.x64.1.5G.iso\casper\filesystem.squashfs\home\hddsuperclone\
hddsuperclone.pdf 299,830 bytes

The weird URI, is I use 7ZIP on Windows as my Swiss Army Knife for computers.
I drill into the ISO. There is a casper directory at the top. it contains "filesystem.squashfs", I do an open inside on that, then there is a home directory, a hddsuperclone directory, and the PDF is in that directory.
(You can open the PDF once the DVD is booted, but it's less convenient
to be using the manual, without being able to read the PDF before you start.
I was reading the PDF on my daily driver computer, while the Test Machine
is doing the clone test.)

Here is the product running right now, to attempt to recover my disk.
While ddrescue quoted something like 40 days to do a recovery,
this product is claiming the operation just might finish today
some time. Never take any of these estimates as being remotely
indicative -- I don't think it will really finish today.

[Picture]

https://i.postimg.cc/pdy7pFcP/HDDSuper-Clone-02.gif

In the picture, you'll notice I did a "swapoff -a" at the start. That's
because my source drive, one of the partitions is a swap partition, and
of course Xubuntu tries to mount all the swap partitions. And while
the machine has enough RAM a swap should not happen, I disabled the
swap so the source drive would not be "polluted" by any swap activity.

The monitor window in the picture, says

"FREE PRO MODE ACTIVE"

And one of the aspects of the PRO mode, is faster access to the sick disk
via the custom driver. The author of the program is giving away the
PRO mode, by loading the license key into it.

My only regret about my source drive, is that the content on it isn't
very "interesting", so it may be difficult to determine the quality
of recovery. There is only one partition on it at 35GB, with presumably
some Linux test distro on it. But when drives fail, we can't be picky
about the contents.

Enjoy,
Paul

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Fri, 28 Nov 2025 11:52:19 -0500, Paul wrote:

I found another program that claims to be a cloning tool. It is like ddrescue, except it uses a separate driver for the "bad" disk, that
driver is capable of keeping the bad driver operating, it truncates
error recovery on the drive and so on.

You probably donrCOt want to use it. The point of ddrescue is to
absolutely minimize further wear and tear on the dying drive, not to
keep running it.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Fri, 11/28/2025 4:20 PM, Lawrence DrCOOliveiro wrote:

On Fri, 28 Nov 2025 11:52:19 -0500, Paul wrote:

I found another program that claims to be a cloning tool. It is like
ddrescue, except it uses a separate driver for the "bad" disk, that
driver is capable of keeping the bad driver operating, it truncates
error recovery on the drive and so on.

You probably donrCOt want to use it. The point of ddrescue is to
absolutely minimize further wear and tear on the dying drive, not to
keep running it.

It takes time to extract all the data.

I was promised that maybe "ddrescue" would take 40 days
for the initial pass.

The competing tool HDDSuperClone is finished now,
and it also can export a ddrescue.log compatible file to
account for the level of progress. The log
currently has around 10,000 lines of text.
The ddrescue, when I stopped it, had 1424 lines
of text.

If I want, I can now go back to ddrescue, and
run the "second pass" to whittle down the 10,000 lines.
I can use the compatible file to do it.

The 250GB sick disk, the data is not important on it, so
I can test potential tools and see what kind of
result they turn up.

ddrescue has no way to reset the disk, when the disk
pulls one of its stunts. The HDDSuperClone had no
problem keeping the sick disk in operational condition.
If the disk pulled a disappearing act, the driver
restarted it (it has "soft reset" and "hard reset"
command sequences of some sort). When I switch back
to ddrescue, I will have to be careful to not ask
it to do things that trigger a controller-initiated reset.
Using --reverse was what I tried before, and it seemed
to stay up longer while being scanned in reverse.
That's a problem specific to this particular drive.

Using a hex editor, I stepped out to the sectors
just before 250,059,350,016 and I found the GPT secondary table
there. That helps me verify I'm at the end of the drive
near that address (that's on the clone drive, not
the sick drive).

Apparently the data recovery community has a "hardware cloner"
that does a disk to disk copy. And it does the same sorts
of things that the HDDSuperClone does with its driver. They're
all using ATA command set to advantage. Whereas ddrescue doesn't
work at that level, so it is harder to tame the sick drive.

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

On Fri, 11/28/2025 4:20 PM, Lawrence DrCOOliveiro wrote:

On Fri, 28 Nov 2025 11:52:19 -0500, Paul wrote:

I found another program that claims to be a cloning tool. It is like
ddrescue, except it uses a separate driver for the "bad" disk, that
driver is capable of keeping the bad driver operating, it truncates
error recovery on the drive and so on.

You probably donrCOt want to use it. The point of ddrescue is to
absolutely minimize further wear and tear on the dying drive, not to
keep running it.

Here is a comparison of what the two programs managed to
do, within the same time period. The one on the right
can still have a secondary ddrescue run done on it, if
I wanted to try.

[Picture] Download Original for best results

https://i.postimg.cc/Zqmzhbhq/clone-recovery-comparison.gif

Paul
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Paul wrote:

The competing tool HDDSuperClone is finished now,

Its site and availability sez:

Up until now, gnu ddrescue has been widely known to be the best free
hard drive cloning tool for failing hard drives. It is open source
and cross platform. But being cross platform has some limitations,
as there are some specialized ways for Linux to send commands to a
drive that have some advantages over standard techniques. So I would
like to introduce HDDSuperClone, which is not open source and only
works on Linux. There is a free version, and also a more advanced
PRO version available for purchase.

... and briefly describes the difference of Pro vs free.

Besides several apps incl hddsuperclone, he provides a live .iso:

The Live CD is now based on Xubuntu. At this time the 32 bit (x86)
version is based on Xubuntu 14.04.5. The 64 bit (x64) version is
based on Xubuntu 18.04.5. Due to the size, the only place I can
currently host the files is on Google Drive.

They are dated '22 Jul of sizes 900 MB & 1.4GB & checksums.
--
Mike Easter
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: alt.os.linux.mint

Mike Easter wrote:

There is a free version, and also a more advanced PRO version
available for purchase.

... and briefly describes the difference of Pro vs free.

Pro is now free:

I no longer have the time to maintain this project. I am planning on
ending major support for HDDSuperClone, so I am making the PRO
version free. I has also been made open source, and the source code
is in the downloads area. The most recent version 2.3.3 downloads
now have the full pro features unlocked without the need for a
license.

--
Mike Easter
--- Synchronet 3.21a-Linux NewsLink 1.2