HD light polling

[faeychild]

Since replacing my failing data drive yesterday I now get a HD light
every half second

Something seems to be polling it.

The accompanying tick noise is coming from the new spinning rust

I have checked the plugs/sockets and don't seem to have left anything
disconnected, not even the optical drive

The polling starts after boot on the login page

The polling runs for "junk" account also

the polling stops at shutdown when the kernel messages start

Where to look?

regards

[William Unruh]

Since we have absolutely no idea what you have on that drive, it is very
hard to say. AFAIK that light will come on on any access of the drive.

Post by faeychild
Since replacing my failing data drive yesterday I now get a HD light
every half second
Something seems to be polling it.
The accompanying tick noise is coming from the new spinning rust
I have checked the plugs/sockets and don't seem to have left anything
disconnected, not even the optical drive
The polling starts after boot on the login page
The polling runs for "junk" account also
the polling stops at shutdown when the kernel messages start
Where to look?
regards

[faeychild]

Post by William Unruh
Since we have absolutely no idea what you have on that drive, it is very
hard to say. AFAIK that light will come on on any access of the drive.

The new drive contains all the data from the old drive, mostly video files
With no access to the data drive and on the login page the drive is
still being polled.
I have also re run lm sensors. no change

[Aragorn]

Post by faeychild

Post by William Unruh
Since we have absolutely no idea what you have on that drive, it is
very hard to say. AFAIK that light will come on on any access of
the drive.

The new drive contains all the data from the old drive, mostly video
files With no access to the data drive and on the login page the
drive is still being polled.

I take it that whatever volume is on the drive, it's mounted at boot
from within /etc/fstab?

Post by faeychild
I have also re run lm sensors. no change

Assuming that you're running Plasma, could it be Baloo that's indexing
the drive? If you don't want that, open up System Settings and navigate
to "Workspace", and then "Search".

You can add a directory path there that must NOT be indexed, or you can
disable indexing altogether, or you can have it index only the filenames
but not the content, and so on.

On my own system here, I've excluded certain directories (and their
subdirs) from indexing, and I'm only indexing filenames, not file
content.

[David W. Hodgins]

Post by faeychild

Post by William Unruh
Since we have absolutely no idea what you have on that drive, it is very
hard to say. AFAIK that light will come on on any access of the drive.

The new drive contains all the data from the old drive, mostly video files
With no access to the data drive and on the login page the drive is
still being polled.
I have also re run lm sensors. no change

The light indicates activity for the controller, not necessarily the hard drive.
Do you have an optical drive on that controller? Automount software will check for
a cd/dvd to automount every 5 seconds.

There are various things that can check, such as udev and software for each
of the main desktop environments.

In kde, run systemsettings5. In the Hardware/Removable storage section, under
Removable Devices, make sure "Enable automounting of removable media" is not
selected.

There are similar settings for Gnome and Xfce, though I don't have the exact
setting locations handy right now.

Regards, Dave Hodgins

[David W. Hodgins]

Post by faeychild

Post by William Unruh
Since we have absolutely no idea what you have on that drive, it is very
hard to say. AFAIK that light will come on on any access of the drive.

The new drive contains all the data from the old drive, mostly video files
With no access to the data drive and on the login page the drive is
still being polled.
I have also re run lm sensors. no change

I just checked. We've had this discussion before ...
https://groups.google.com/g/alt.os.linux.mandriva/c/dc5JE28sK9M/m/nXg3IsbWjYsJ?pli=1

Regards, Dave Hodgins

[William Unruh]

Post by faeychild

Post by William Unruh
Since we have absolutely no idea what you have on that drive, it is very
hard to say. AFAIK that light will come on on any access of the drive.

The new drive contains all the data from the old drive, mostly video files
With no access to the data drive and on the login page the drive is
still being polled.

What does "no access to the data drive" mean? Did you unplug the drive.

Post by faeychild
I have also re run lm sensors. no change

[faeychild]

On 28/5/22 16:56, faeychild wrote:


I will relate the sad tale

I had a slowing machine. The responsiveness , file loading, desktop
actions sometimes lagged.

The primary drive with OS installed is a SDD. Second drive is spinning
rust, Data only mostly video files

Fsck passed the HD partition and smartctl found some sectors disabled.

Other than that the computer was fine, nothing strange going on.


I replaced the HD and copied all data across from the old drive.

The system was back to being zippy. All good



Then I noticed the new "functions. "

The HD light was flashing regularly. I could hear the new HD tick. I
thought, It's lust logging. But It would also tick after boot on the
login screen. The same for "junk" user. I was skeptical to find the
data HD accessed at the login screen, but something is
This ran on yesterday until late last night when I noticed the flashing
had ceased. To be replaced this morning by a regular soft thump from
the drive with no HD light activation


The second new function is
USB flash memory is no longer mounted or showing up in Dolphin

When the flash memory is inserted it is listed in dmesg and listed in
devices in gparted But it is no longer auto-mounted.

Is this co-incident with the new HD or just bad timing (rhet)

Should I take a week off and drink :-)

And to David. Yep I do remember the old problem with HAL polling the
optical drive

regards

[faeychild]

On 29/5/22 10:32, faeychild wrote:

After failing all morning I tried the USB stick again
It was recognized and auto mounted and popped up in Dolphin
Three times flawless

Actually to be precise, the stick is not auto mounted. It is just
recognized and listed in Dolphin. It has a right click menu with mount
option. A Left click performs the same function



I would also suspect the USB extension cable but the USB stick didn't
work yesterday when plugged directly into the socket. All sticks failed
not just one

I'm seriously considering the Drinking

[David W. Hodgins]

Post by faeychild
After failing all morning I tried the USB stick again
It was recognized and auto mounted and popped up in Dolphin
Three times flawless

Glad it's working now.

Post by faeychild
Actually to be precise, the stick is not auto mounted. It is just
recognized and listed in Dolphin. It has a right click menu with mount
option. A Left click performs the same function
I would also suspect the USB extension cable but the USB stick didn't
work yesterday when plugged directly into the socket. All sticks failed
not just one
I'm seriously considering the Drinking

Lol.

The optical drive polling starts before the login screen is presented. While
there is no notifying the user, since there are no logged in users, the polling
starts when udev starts, or haldaemon for very old installs.

Regarding the the usb not working/working, your guess is as good as mine.

Regards

[TJ]

Post by David W. Hodgins
Regarding the the usb not working/working, your guess is as good as mine.

My first suspect would be a worn usb port, where the user has to plug in
the stick just so or it doesn't work. I have a couple like that on an
old case that I've had for many years.

My solution was to buy a four-port hub that fits into the floppy bay,
and connect it to an internal header.

TJ

[faeychild]

Post by David W. Hodgins
Regarding the the usb not working/working, your guess is as good as mine.
Regards

Eventually, it's always a guess


The new HD is making a regular quiet clunk with no accompanying light


I'm guessing maybe a faulty drive after all, or not
I suppose a new drive could be defective and difficult to prove

Or perhaps I'm unwittingly living on an old indigenous burial ground

regards

[David W. Hodgins]

Post by faeychild
The new HD is making a regular quiet clunk with no accompanying light

Parking the heads? See the -J option of "man hdparm".

Regards, Dave Hodgins

[faeychild]

On Sun, 29 May 2022 17:47:42 -0400, faeychild

Post by faeychild
The new HD is making a regular quiet clunk with no accompanying light

Parking the heads? See the -J option of "man hdparm".
Regards, Dave Hodgins

That's a disappointing read and changing it with Windows exe file will
be difficult

"A full power cycle is required for any change in setting to take effect"

I suppose this means turn it off and on again.

Guess who wont buy WD next time

regards

[faeychild]

On 31/5/22 08:43, faeychild wrote:


OK If you enter WDIDLE3.EXE in the search function at
https://www.westerndigital.com.

The results do not inspire much confidence
Five complaints are listed. One in Russian

I bought a lemon, didn't I?

[faeychild]

OK  If you enter WDIDLE3.EXE in the search function at
https://www.westerndigital.com.
The results do not inspire much confidence
Five complaints are listed. One in Russian
I bought a lemon, didn't I?

There be hope yet

I found


"i used the alternative opensource (actually FreeSoftware) linux based
tool --> http://idle3-tools.sourceforge.net/"


It must be compiled first. And a first compile for me!!
That will require aa quiet afternoon

regards

[Aragorn]

Post by faeychild
Guess who wont buy WD next time

Western Digital in and of itself is okay. What matters is the
type/model of HDD you buy.

Best is to buy a model from the Black or Enterprise series, and WD Green
is a model range you had better stay away from, because they park and
spin themselves down like crazy, and they don't survive for very long,
exactly because of that.

Besides, it's pointless anyway. The electricity you save by spinning
the drive down is all wasted again when it has to spin back up, which
requires a lot more effort from the spindle motor and draws far more
current from the PSU than just keeping the drive spinning all the time.

Furthermore, the heads of a HDD are kept afloat above the platters by
the principle of aerodynamics — the air inside the drive housing is
moved by the spinning of the platters. This means that when the
platters aren't spinning — or aren't spinning fast enough — the heads
will be in contact with the platter surface.

This is why the heads need to be parked. In some HDD designs — e.g.
the IBM-designed UltraStor range that was later sold to Hitachi — the
heads are parked on a ramp outside of the platter circumference, but in
other designs — most Seagate and Quantum designs — the heads are
parked on the innermost cylinder of the platters.

This innermost cylinder will then typically not be coated with the
ferromagnetic material that the rest of the platters have, but will
rather have a smooth surface. However, this method of parking in turn
means that the heads will then still be in contact with the platter
surface for a while when the platters start rotating, and that they
will also already be in contact with the platter surface again due to
the loss of aerodynamic lift before the spindle motor has fully spooled
down. This friction generates extra wear on the heads, and possibly
also on the actuators, because they will experience a lateral force
before the heads can clear the platters.

If you want your HDDs to survive for a long time, keep them spinning.
Best advice I can give you. ;)

[faeychild]

Post by Aragorn
If you want your HDDs to survive for a long time, keep them spinning.
Best advice I can give you. ;)

I can only agree with you, Aragorn.

My first thought was why park the heads every eight seconds on an
operating HD.
It seems like a fairly stupid decision, maybe management one :-)


My drive is thumping about every four seconds. I will keep an eye on
it for a week and think about trying the utility

The HD is a Western Digital WD Red Pro 6TB WD6003FFBX 3.5in NAS Hard Drive.

I imagine that continually parking the heads on a NAS drive would
severely impact performance

regards

[red floyd]

Post by Aragorn
If you want your HDDs to survive for a long time, keep them spinning.
Best advice I can give you. ;)

 I can only agree with you, Aragorn.
My first thought was why park the heads every eight seconds on an
operating HD.
It seems like a fairly stupid decision, maybe management one :-)
 My drive is thumping about every four seconds.  I will keep an eye on
it for a week and think about trying the utility
The HD is a  Western Digital WD Red Pro 6TB WD6003FFBX 3.5in NAS Hard
Drive.
I imagine that continually parking the heads on a NAS drive would
severely impact performance

I'm using Red Pros (10GB) in my NAS as well. They don't seem to thump
every 4 seconds. I'm wondering if you have a bad drive.

[faeychild]

I'm using Red Pros (10GB) in my NAS as well.  They don't seem to thump
every 4 seconds.   I'm wondering if you have a bad drive.

It is a new drive several days old and that though is in the back of my
mind also but proving it to myself and then the vendor would would be
difficult

I may try the source forge tool sometime and check if the thump
timing alters

regards

[David W. Hodgins]

Post by faeychild
It is a new drive several days old and that though is in the back of my
mind also but proving it to myself and then the vendor would would be
difficult
I may try the source forge tool sometime and check if the thump
timing alters

Have you tried it in linux, even if it's just from a live iso?

In linux try (as root" "hdparm -J /dev/sd?" (with the correct drive letter)
to see what the current value is, and then "hdparm -J 30 /dev/sd?" to
change it. As it requires a full poweroff/on to take effect, the change should
work in both linux and windows.

Regards, Dave Hodgins

[faeychild]

Post by David W. Hodgins
In linux try (as root" "hdparm -J /dev/sd?" (with the correct drive letter)
to see what the current value is, and then "hdparm -J 30 /dev/sd?" to
change it. As it requires a full poweroff/on to take effect, the change should
work in both linux and windows.
Regards, Dave Hodgins

Not what I expected

[***@unimatrix ~]# hdparm -J /dev/sda

/dev/sda:
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
wdidle3 = disabled
[***@unimatrix ~]#

[David W. Hodgins]

Post by faeychild

Post by David W. Hodgins
In linux try (as root" "hdparm -J /dev/sd?" (with the correct drive letter)
to see what the current value is, and then "hdparm -J 30 /dev/sd?" to
change it. As it requires a full poweroff/on to take effect, the change should
work in both linux and windows.
Regards, Dave Hodgins

Not what I expected
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
00 21 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
wdidle3 = disabled

Is this drive connected over a usb interface? It needs to be connected directly
to a sata controller.

Regards, Dave Hodgins

[faeychild]

Post by David W. Hodgins
Is this drive connected over a usb interface? It needs to be connected directly
to a sata controller.
Regards, Dave Hodgins

It is connected to a sata port on the motherboard


System Information
Manufacturer: Gigabyte Technology Co., Ltd.
Product Name: B250M-D3H



ls -l /dev/disk/by-path/
total 0
lrwxrwxrwx 1 root root 9 Jun 2 12:52 pci-0000:00:17.0-ata-2 -> ../../sda
lrwxrwxrwx 1 root root 9 Jun 2 12:52 pci-0000:00:17.0-ata-2.0 -> ../../sda
lrwxrwxrwx 1 root root 10 Jun 2 12:52 pci-0000:00:17.0-ata-2.0-part1 ->
../../sda1
lrwxrwxrwx 1 root root 10 Jun 2 12:52 pci-0000:00:17.0-ata-2-part1 ->
../../sda1
lrwxrwxrwx 1 root root 9 Jun 2 12:49 pci-0000:00:17.0-ata-4 -> ../../sr0
lrwxrwxrwx 1 root root 9 Jun 2 12:49 pci-0000:00:17.0-ata-4.0 -> ../../sr0
lrwxrwxrwx 1 root root 13 Jun 2 12:49 pci-0000:02:00.0-nvme-1 ->
../../nvme0n1


I have run systemrescue from a stick and the drive is still thumping

I also pulled the sata cable from the drive and it still thumps.

Maybe it's not looking good

regards

[Jim]

Post by faeychild

Post by David W. Hodgins
In linux try (as root" "hdparm -J /dev/sd?" (with the correct drive letter)
to see what the current value is, and then "hdparm -J 30 /dev/sd?" to
change it. As it requires a full poweroff/on to take effect, the change
should work in both linux and windows.
Regards, Dave Hodgins

Not what I expected
SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00
bad/missing sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00 00 21 04
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 SG_IO: bad/missing
sense data, sb[]: 70 00 05 00 00 00 00 0a 04 53 00 00 21 04 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 SG_IO: bad/missing sense data,
sb[]: 70 00 05 00 00 00 00 0a 04 53 00 00 21 04 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00
wdidle3 = disabled

Replace 53 in the above each time with 51 and you get exactly what
I get when I run hdparm -J /dev/sda (or sdb) on my main machine.

My two drives are Samsung SSD 750 EVO solid state drives. No head
parking involved.

My guess is that a connection or cable is bad.

Cheers!

jim b.