bad hard disk noise on shutdown

Nobody else can confirm this? I noticed also on other laptops with SATA hard-disk. I'm concerned about the health of my hard-disk...

Happens also on my Dell Inspiron 9400. Running Kubuntu 6.06 with no problem on the shutdown.

Tried fully updated kubuntu edgy => and the strange noise is there
Tried kubuntu feisty => and the strange noise is there

This was classified as a kernal bug, but edgy and feisty run different kernels. Looks to me as a teardown problem.

Re-confirm against 2.6.20, please.

Confirmed on 2.6.20-2.2 on my Acer and Compaq and several duplicates.

Confirmed on 2.6.20-2-generic on the hp dv2172ea (dv2000 series) notebook.

This should be marked as CRITICAL imho. I care of the health of my hardware!

One of the duplicates was already marked of importance "high", so it's just a question of someone with the right permissions to set it here as well. If it can cause hardware failure over time, it should probably be "critical"...

confirmed also on dell inspiron 6400 laptop

Can this be retested against 2.6.20-3 please?

I can confirm this too on a Acer TM 8000, but with 2.6.19 kernel(2.6.20 doesn't boot on my laptop)

It doesn't occurs it I use sysrq key to emergency reboot

Actually this also happened with magic keys...

Tested with 2.6.20-3: the bug is still present.
Bye
Marco

I've rebuild all initramfs images with update-initramfs -u -k all (initramfs-tools version 0.85eubuntu1).

the HD heads are parked now with 2.6.19-7 kernel only when shutting down but not when rebooting.

Same thing with 2.6.20 but I cannot be 100% sure because my laptop is rebooting instead of shutting down (bug #71040). But just before the laptop should be shut down, I heard the HD heads be parked.

It didn't do the trick for me.

Same bad noise on 2.6.20-4.6. I also have bug #71040 since 2.6.20, but it screams out before it reboots.

Bug still present even with 2.6.20-5 kernel.
Bye
Marco

Confirmed on kernel 2.6.17-10-generic.

What is the solution to this problem?? Help needed !

Hi Mehul Dixit,
There is no solution yet. The problem is found in other Linux Distro's to so it is marked as a kernel bug.

I hope there wil be a solution soon, because this keeps me from going to a newer Kubuntu. I'm on 6.06 LTS.

bye More

Hi All,

I just tested opensuse 10.2 with kernel-default-2.6.18.2-34.i586.rpm on my laptop and it does a very nice shutdown without any strange hard disk noise.

So I don't think it's a linux kernel problem, but a (k)ubuntu problem.

What could cause this? Maybe the new shutdown methode "teardown"?
Kubuntu 6.06 LTS => ok => doesn't use teardown
Kubuntu 6.10 => not ok => uses teardown
Kubuntu 7.04 beta => not ok => uses teardown
Opensuse 10.2 => ok => doesn't use teardown.

So maybe it is teardown????
Bye More

Same problem on 2.6.20-6.11.

Hi Tormod,

Could you, if possible, test OpenSuse 10.2? If you could confirm that OpenSuse 10.2 isn't making a strange shutdown noise we would be 1 step closer to a solution.

Thanks, More

Does OpenSuse have a net-install or something minimal I can test without downloading a full CD? I think it's a question of which disk drivers are being used, ide-generic, libata or whatever they are called.

As mentioned earlier, I have bug #71040 as well, which makes it reboot when I choose shutdown. If I choose Reboot there is no noise, but if I choose Shutdown I get the noise, even though the reboot seems to shut off the laptop completely for a second or two before starting again. So for me the drive-friendly way to shut down is to reboot, enter grub command mode and enter halt, then turn off.

Hi Tormod,

OpenSuse has an ftp install option. I never tried that option, but here is the docu http://en.opensuse.org/Installation_without_CD

I'm not having bug #71040 problems. Hmmm that reboot and then enter grub command mode and enter halt, then turn off I must try.

thanks for testing. I will test opensuse again and have a look on what hard disk driver gets loaded.

Thanks, More

Strangely halt does not work on my grub. Is this a kernel bug too? I am having to reboot into windows everytime & then shutdown my comp.

Halt (in grub) doesn't work for me either, that is, the machine is halted but not turned off. That's probably a grub bug. I have to keep the power button pressed to turn it off, but that's safe once the machine is halted and the hard drive has parked.

Happens to me too. It's weird because I've installed the same Ubuntu on my other laptop and it shuts down just fine.
I'm running an Asus Z63a (Pentium-M 1.73 Ghz 533 MHz FSB, 60Gb Samsung Spinpoint M HDD)
The other laptop that works fine is an Asus M6Ne (Pentium-M 1.5 GHz 400 MHz FSB, 80Gb Toshiba HDD)

So it's not a special brand of harddisk that is giving us these troubles. My Dell uses a Hitachi 80GB disk

I can only confirm; the same happens since upgrade from dapper to edgy (using aptitude)
I use a Dell inspiron 9400 with SATA harddisk 160Gb (don't ask me about brand). Windows on the same machine still shuts down properly.

would there be a way to turn off this teardown feature? It would at the same time confirm if this is the problem...

Teardown is just a specification on how to stop processes the most economic way, and is not related to this.

To easily compare between kernels, boot with the kernel option:
init=/bin/bash
(and take "quiet" and "splash" away). Then turn off the machine with "halt -f -p".

My Compaq (with the via chipset) turns off quietly now, I think it has to do with the return to ide/pci/via82cxxx.ko from the ata/pata_via adventure.

The upstream bug has a link to this interesting discussion: http://mailman.linux-thinkpad.org/pipermail/linux-thinkpad/2007-January/037903.html which again points to this: http://<email address hidden>/msg02960.html

What about hibernate? Do you guys hear the click when entering hibernate?
I sure do... rebooting and suspend-to-RAM work fine though.

I'm by no means an expert when it comes to this stuff... but by skimming through the link that Tormod posted, it appears that there's in a bug in the kernel(s) that have to do with our hardware being recognized as SCSI? (i.e /dev/sd*), which makes sense because my partitions are recognized as sda1 - sda4, whereas on my other machine that doesn't have the shutdown noise, the partitions are recognized as hda instead of sda.

Is it possible to produce a version of the above patch that apply to kernel 2.6.17 of ubuntu 6.10?
However, i leaved a post on lmkl, http://groups.google.com/group/linux.kernel/browse_frm/thread/a08eb3f2887ae852/fa787ed42c1e3097?lnk=gst&q=weird&rnum=9 , regarding this problem. Please, if you are interested in seeing this bug resolved asap, put some contributions to it, as it seems very difficult to gain attention on this issue and, in my opinion, it's yet to be determined if this is a regression of a lack of feature. For example, can someone affirm that this bug wasn't present in ubuntu dapper drake using libata drivers?

If we have an IDE drive, is there a way to NOT use libata and to use the old IDE driver (without having to reinstall)? Perhaps this could serve as a temporary fix for those with IDE drives instead of SCSI/SATA drives

Listing libata in /etc/modprobe.d/blacklist should prevent it from being loaded, and as such, fall back to the old IDE code.

I'm not exactly a Linux guru, so I'm not exactly sure how I'd blacklist libata. I tried googling to find out how, but nothing turned up. I've tried editing /etc/modprobe.d/blacklist with gedit but I don't know what to add to blacklist it; Is it as simple as writing

blacklist libata

? I believe I've tried that and it didn't work.

Any help would be appreciated!

Thanks

Okay, I tried blacklisting libata and then running the command

> update-initramfs -u

But when I rebooted the OS, I didn't see any changes. IIRC, using the old IDE driver should make my partitions be recognized by "hdx" rather than "sdx", right? Anyhow, I still see my partitions as "sda" and I still hear the noise when shutting down.

Any other ways to disable libata without recompiling the kernel?

I think blacklisting doesn't work for the modules in the initrd (as commented in bug #72765), and libata will get loaded anyway. You can try removing the libata module manually from your initrd.gz, see https://wiki.ubuntu.com/CustomizeLiveInitrd for how to modify an initrd.

I'm kinda overwhelmed with all of this tweaking (like I said, I'm not exactly a Linux guru...), so I'm going to install Dapper later today and see if the same thing happens.

Oddly, no one except few people are complaining about this bug in the RIGHT places, linux kernel bugzilla and lkml. What do you need? A f***ing workaround that is good for 2 persons in the planet or a REAL solution???

Send additional details on http://bugzilla.kernel.org/show_bug.cgi?id=7674 or just leave a message at <email address hidden> (subscription is not needed and you can follow it from http://groups.google.com/group/linux.kernel/topics) asking for kernel bug #7674.

Is there anything new with this bug? I'm asking because it's an absolute showstopper for me; My machine is relegated to XP until something gets figured out...I'm too worried about damaging my HD. I guess I was hoping that some ubuntu-geek-genius had found a way to solve it.....

It seems, though, I suppose, from what I'm reading, that this isn't a solvable problem from the ubuntu end of things, until something comes out of the main kernel development folks? Or do I misunderstand the process?

I filed a bug there under kernel bugzilla #7674 as well, but I can't find anything updated or remotely new there either.

you can still use the 2.6.15 kernel; works fine on my edgy and shutdown is done properly.

Dan, if you can reboot without getting the noise, you can halt the machine from grub, as explained in a previous comment. Inconvenient, but safe.

I was promised that a fix will eventually arrive. It would take longer because a code refactory of the libata layer was in program.

Patch for stop_on_shutdown was added. This bug isn't very critical, IMO. The bug is just annoying, there's no danger to the drives (no matter how the thing sounds when it stops).

The noise that is being described sounds just like a disk-head being parked (these are the same laptop drives that say "drive may rattle, this is normal" on the label), and I'd be more worried if it /wasn't/ hasn't the disk-heads. It's quite satisfying to know that the heads have been 'clicked' into place.

Tried 2.6.20-13-generic (which has the stop_on_shutdown) on an Acer TravelMate 8101, and it still makes that ugly, loud sound at shutdown. BTW the sound has nothing to do with the little rattle that comes when the disk hasn't been used for a while (for instance after halting in grub), which I though was the "parking" sound. I have tried to record it, but my recording gear is not good enough.

Fixed in mainstream http://article.gmane.org/gmane.linux.scsi/30489 . Lot of infrastructure here, probably 2.6.22 stuff? Please note that the FULL fix will need shutdown(8) modifies, as described here http://article.gmane.org/gmane.linux.ide/17392 .

In the same message is described a meantime workaround, that consist of a temporary setting (already scheduled for removal) in the libata module. Here the extract:

> This patch implements module parameter libata.spindown_compat which,
> when set to one (default value), prevents libata from spinning down
> disks on shutdown thus avoiding double spinning down. Note that
> libata spins down disks for suspend to mem and disk, so with
> libata.spindown_compat set to one, disks should be properly spun down
> in all cases without modifying shutdown(8).
>
> shutdown(8) should be fixed eventually tho. Some drive do spin up on
> SYNCHRONZE_CACHE even when their cache is clean. Those disks
> currently spin up briefly when sd tries to shutdown the device and
> then the machine powers off immediately, which can't be good for the
> head. We can't skip SYNCHRONIZE_CACHE during shudown as it can be
> dangerous data integrity-wise.
>
> So, this spindown_compat parameter is already scheduled for removal by
> the end of the next year and here's what shutdown(8) should do.
>
> 1. Check whether /sys/modules/libata/parameters/spindown_compat
> exists. If it does, write 0 to it.
>
> For each libata harddisk {
>
> 2. Check whether /sys/class/scsi_disk/h:c:i:l/manage_start_stop
> exists. If so, write 1 to it and continue; otherwise, fall
> through to #3.
>
> 3. Synchronize cache and spin down as before.
>
> }

If i can ask, is there a mainstream for shutdown(8) or it vary between distros?

Just heard that Ubuntu Feisty beta is available. Is the bug fixed in that?

Are we in a holding pattern on this patch (i.e., are we waiting for the next kernel release) or do you know? I've been updating feisty all the way to this morning, and still no change. Wonder if this is going to come out on 2.6.22 or something?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Daniel R. Gilliam
2426 Stone Post Terrace
Richmond, Virginia 23233
(804)364-1153 (h)
(804)690-1407 (c)
<email address hidden>
**************************
"Give me a stopwatch and a map, and I'll fly the Alps in a plane with no windows." --Mr. Kamarov, "The Hunt for Red October"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I reopen the bug, because nothing has changed on my Acer, unfortunately not with 2.6.20-14-generic either.

kernel: 2.6.20-14-generic / Inspiron 8600

Bug is still present!

Still present:

2.6.20-14-generic #2 SMP i686 GNU/Linux

(Samsung P35 - HHD Samsung HM100JC 100gb)

Same problem here.

FS V2030 with Samsung SATA drive HM040HI.

Kubuntu Edgy 6.10 kernel 2.6.17-11-generic.

Still the same with mine as well: Toshiba Satellite M55-S3314 on 2.6.20-14-generic

Same for me on my Samsung X11 (HD: HM100JC)

I managed it with a script in /etc/rc0.d/ which writes "1" to "/sys/class/scsi_disk/0:0:0:0/stop_on_shutdown" but this should only be a temporary workaround

Martin, thanks a lot for that tip! It shuts down nice and quietly now, using that /sys setting.

Martin or Tormod, can you guys e-mail that script to me? I'm still pretty new to Linux and I'm not exactly sure how I would go about writing a script to do such a command.
You can reach me at <email address hidden>

Thanks

Simply write a file at "/etc/rc0.d/S00hdd-shutdown-workaround" which includes this two lines:

#!/bin/sh
echo 1 > /sys/class/scsi_disk/0\:0\:0\:0/stop_on_shutdown

If you have more than one directory at "/sys/class/scsi_disk/" try to find out which of it is your internal harddisk (e.g. "cat /sys/class/scsi_disk/a:b:c:d/device/model") If you have more than one harddisk add another line for each of them.

Then make it executable ("chmod +x filename")
At the next shutdown it should park the hd heads.

Thanks, Martin.
I created the script... but stupid me, I'm running Ubuntu Edgy which doesn't have the stop_on_shutdown patch implemented.
Is it possible for someone to send me the patch, or do I have to upgrade the kernel?

Thanks a lot Martin.

The problem seems resolved.

Will these be merged to Ubuntu kernel:
http://article.gmane.org/gmane.linux.scsi/30489
http://article.gmane.org/gmane.linux.ide/17392

BTW i'm having this problem on Fujitsu-Siemens Amilo M7440G

This problem also exists on macbook C2Ds

Hi all,
The script from Martin Koßler works nice, but doesn't work on a Hibernate action.

A better working solution for all runlevels, Hibernate and Suspend is to place the script in "/ect/init.d/hdd-shutdown-workaround" and softlink it to /etc/rcS.d/S99hdd-shutdown-workaround with the command: ln -s /etc/init.d/hdd-shutdown-workaround /etc/rcS.d/S99hdd-shutdown-workaround
Do this as root user.
This will set the noise less shutdown during startup and then all differtent shutdown options will work fine.

More

We (libata kernel dudes) are working on a simpler kernel level fix for this now, which may not require any immediate changes to scripts.

If anyone here can patch/rebuild kernels, then I have some code that needs testing before we queue it into 2.6.21, and then make it available for the Ubuntu kernel folks. Email me for the patch if you can retest with it more or less immediately.

Thanks.

<email address hidden>

Oh, forgot we can do attachments here (duh!).

This is one patch that's being considered, which should get rid of the pop.
Can anyone here who has the problem please try this immediately and report back?

Thanks

> please try this immediately and report back?

Which version of the kernel this patch apply?

Francesco Pretto wrote:
>> please try this immediately and report back?
>
> Which version of the kernel this patch apply?

Anything from about 2.6.18 onwards.

Cheers
--
Mark Lord
Real-Time Remedies Inc.
<email address hidden>

Mark Lord wrote:
> Francesco Pretto wrote:
>>> please try this immediately and report back?
>> Which version of the kernel this patch apply?
>
> Anything from about 2.6.18 onwards.

Early reports indicate that this patch likely does NOT solve the problem.

Cheers
--
Mark Lord
Real-Time Remedies Inc.
<email address hidden>

>Early reports indicate that this patch likely does NOT solve the problem.

Did not work for me on a fresh Ubuntu Feisty Fawn install but I'm sure you*ll make it work soon. :-)

> Early reports indicate that this patch likely does NOT solve the problem.

The fact is that this is not a bug/regression of the kernel. It's a lack of feature.
The early original patch here, http://www.nabble.com/(fwd)--PATCH--sd:-implement-stop_on_shutdown-t3049703.html , that was already merged, need mandatory changes in userspace, so changes to upstart shutdown(8) or workaround as described in some posts here. The definitive patches here:

http://article.gmane.org/gmane.linux.scsi/30489
http://article.gmane.org/gmane.linux.ide/17392

are probably too much invasive to be merged in a stable, distro customized kernel, but don't need changes in userspace (beacuse they use a kernel-space workaround, already scheduled for removal). So, from a distro perspective, there are few options:
1) Keep original early patch and modify upstart shutdown(8) now (i was waiting for the definitive patches to be merged in 2.6.22-rc before submitting a bug to upstart...);
2) Merge the definitive patches;
3) || die

From a user perspective, simply use the early patch (that should be already merged in ubuntu 2.6.20 kernel, right?) plus the workarounds described here by Martin Koßler and More.

I confirm the problem in Feisty Fawn, stock kernel from CD, happens both by hibernating and shutdown.
The HD is a Samsung PATA 80Gb.

I confirm that in Feisty Fawn,using stock kernel.I have 2 WD HDDs..I tried the solution with the script but it only works for one of the disks.How can i do this for both of them?I didn't quite understand it.... :P

Aramil, for more than one disk, try this:
for disk in /sys/class/scsi_disk/* ; do echo 1 > $disk/stop_on_shutdown ; done

(BTW, I personally prefer to add this to /etc/rc.local, instead of creating a new script file)

That works for me, Tormod, thank you. Maybe this script could be added as an intermediate solution, until a fix for the kernel can be released?
It would be great to not have the disks smashed up.

Same problem to. feisty + asus laptop (M6v)... Thre strange situation is that with edgy i didn't have this kind of problem... Even with dapper

Same problem here with. The sound is like shuting down manually the laptop. The laptop has Intel Centrino 1.6G and Hitachi 40G hard drive (SCSI Device - ATA HITACHI_DK23FA-4).

I had no problems with edgy and dapper. If I boot with edgy kernel (2.17) I still have no problems.

I tried the workaround of Martin and More. It didn't work for me.
I made a file in /etc/init.d with the following two lines:

#!/bin/sh
echo 1 > /sys/class/scsi_device/0\:0\:0\:0/stop_on_shutdown

(only had to change "scsi_disk" for "scsi_device")

then

sudo chmod +x filename
sudo ln -s /etc/rc0.d/S99hdd-shutdown-workaround /etc/init.d/hdd-shutdown-workaround

I guessed this didn't work because "S99" meant it was called after S90halt so the computer was shutdown before. so I changed it to "S00hdd-shutdown-workaround" (like Martin wrote) but still nothing changed. I had no noise in Dapper, then noise in Edgy with kernel 2.6.17 but not with kernel 2.6.15, and still the noise in Feisty with kernel 2.6.20.

I don't have this file (or directory?) /sys/class/scsi_disk/0\:0\:0\:0/stop_on_shutdown . only /sys/class/scsi_device/0\:0\:0\:0/
Am I missing something here?

I noticed something weird in phpsysinfo, which I do not know if it has to do with the bug or should I report a new one. With 2.6.17 the hd and cdrom are shown as IDE and with 2.6.20 as SCSI devices. Here are the interesting parts of the phpsysinfo.

Model Intel(R) Pentium(R) M processor 1.60GHz

Kernel Version 2.6.20-15-generic (SMP)

SCSI Devices
- ATA HITACHI_DK23FA-4 (Direct-Access)
- COMPAL TSB24H1 DVD/CDRW (CD-ROM)

Kernel Version 2.6.17-11-generic (SMP)

IDE Devices
- hdc: COMPAL TSB24H1 Slim DVD-ROM/CD-RW Drive
- hda: HITACHI_DK23FA-40 (Capacity: 37.26 GB)

@hetvliegend...,

I'm linking it in rcS.d instead what you are doing in rc0.d

Sxx means start during boot, kxx means kill during shutdown. When putting the script in rsS.d it is always run during startup and this means that the stop_on_shutdown parameter is always set. Doesn't matter if you shutdown, reboot, suspend or hibernate.

More

I have tried More's workaround and it indeed works for shutdown, suspend, and hibernate. (it never made a sound when rebooting)

The definitive patches [1] [2] have been merged in mainstream 2.6.22rc1. I opened the bug #114683 to request changes in upstart shutdown(8).

[1]
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=9666f4009c22f6520ac3fb8a19c9e32ab973e828
[2]
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=920a4b1038e442700a1cfac77ea7e20bd615a2c3

The problem seems to be fixed for me.
With the latest kernel-update I noticed that my drive is mapped to /dev/hda again (instead of sda). So I removed the workaround (which would not effect the drive anyway since it doesnt show up /sys/class/scsi_disk/ anymore) and the hdd still shuts down properly.

Yes, confirmed. The problem is solved with the new kernel upgrade.

NO, unconfirm!!!

Fix isn't working. Still have the hdd making noise when I shutdown the system.

Also the harddisk still shows as /dev/sda.

Martin + Jan what kernel are you running? I'm running Ubuntu 7.04 with kernel 2.6.20-16-generic

What did you do to get /dev/hda back?

Thanks,
More

Just apt-get upgraded to kernel 2.6.20-16-generic

That is what I did this morning. We are on the same kernel.

I'm running a Dell Inspiron 9400 with a Hitachi 80GB harddisk

If the reason it's "fixed" for Jan and Martin is that their disk drivers have been reverted to /dev/hda, I would not qualify it as bug resolved. It will then come back when your disk drivers are /dev/sda again, which they might be in a future kernel. That said, this issue will probably be solved correctly in the 2.6.22 Gutsy kernel. Until then the rest of us with /dev/sda drivers have to live with the above stop_on_shutdown workaround.

Fix doesn't work for everybody.

The problem is solved for me too. The hard drives are mapped at /dev/hda.

My drives are still mapped at /dev/sda. I can't say if the fix has worked or not because I'm reluctant to turn off the stop_on_shutdown workaround, but I'm confident that since my drives are still mapped as sda, it hasn't worked for me either.

Hard drive now mapped to /dev/hda, but the fix didn't work for me.

On 5/29/07, Emilio Pozuelo Monfort <email address hidden> wrote:
> ** Changed in: linux-source-2.6.20 (Ubuntu)
> Status: Rejected => Confirmed
>
> --
> bad hard disk noise on shutdown
> https://bugs.launchpad.net/bugs/67810
> You received this bug notification because you are a direct subscriber
> of the bug.
>

Someone can confirm me that the patch that add "stop_on_shutdown" is still in the latest ubuntu 2.6.20 kernel? Thanks

After the latest update the problem came back again. The devices are mapped at /dev/sda and the bad sound is back again. And I cannot go back to the kernel before the update.

Same for me on my Samsung X11 notebook like Tormod predicted. Does anyone know if theres a possibility to get the driver from the last kernel version again, e.g. through a option in /etc/modules or s.th. like that?

Martin, see comment #39 on that. But isn't your workaround in comment #57 a much easier solution?

Yes if I have to modify the initrd my workaround is much better so i'll restore it in rcS.d.

For those who might assume this is a hardware issue; Ubuntu Edgy or Windows XP do not do this!
Whether its actually causing damage or not it certainly sounds like it is -
This should definately be a critical issue!

this is happening in my edgy install

System: IBM Thinkpad R40.
Hard drive: Hitachi Travelstar 80GN, Model Number: IC25N040ATMR04. It's a PATA drive.

I observed this noise for the first time in Ubuntu 7.04. I had no such problems with 6.06 or 6.10. Martin Koßler's workaround works fine.

I'd like to cite from <http://www.hitachigst.com/tech/techlib.nsf/techdocs/85CC1FF9F3F11FE187256C4F0052E6B6/$file/80GNSpec2.0.pdf>:

--8<---------------cut here---------------start------------->8---
Emergency unload is intended to be invoked in rare situations. Because
this operation is inherently uncontrolled, it is more mechanically
stressful than a normal unload. A single emergency unload operation is
more stressful than 100 normal unloads. Use of emergency unload
reduces the start/stop life of the drive at a rate at least 100 times
faster than that of normal unload and may damage the drive.
--8<---------------cut here---------------end--------------->8---

Is that what happens without Martin's patch? If so, it doesn't sound too good ...

I guess the "Power-Off_Retract_Count" that one gets with "sudo smartctl --device=ata --attributes /dev/sda" tells how often the hard drive has gone through an emergency unload. Can somebody confirm that? The "Power-Off_Retract_Count" of my drive is 14, and I think that number would make sense.

The "Power-Off_Retract_Count" of my drive (SATA) is 358, and considering that I have used Ubuntu 6.10 for nearly 6 months without a proper fix, I think that number makes sense. Thanks Ubuntu, I'll go for SSD next time. ;-P

I confirm this bug on a acer aspire 5601, I'm using feisty 7.04.
~$ uname -a
Linux laptop 2.6.20-16-generic #2 SMP Thu Jun 7 20:19:32 UTC 2007 i686 GNU/Linux

Happens the same with my laptop Dell Inspiron 1501 and using Feisty

~$ uname -a
Linux laptop 2.6.20-16-generic #2 SMP Thu Jun 7 20:19:32 UTC 2007 i686 GNU/Linux

According to <https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=243021#c23>, the "Power-Off_Retract_Count" seems to be the correct "counter" for how often your disk has undergone an emergency unload.

Just to confirm, I have had this issue with all distros using 2.6.20.* and my Hitachi HD. The 2.6.22+ kernel has resolved this issue!! I used it with Fiesty and then upgraded to Gutsy tribe4. I will try Mint and Dream next and if I'm brave enough I'll try some RPM distros too.
I hope this helps most of you if you have not already tried it.

Ubuntu Feisty 7.04 here. Just want to confirm that it makes that clicking + long spinning noise at menu-shutdown in Dell Inspiron 6000, too.

(Inspiron 6000 Dell) I upgraded my kernel to 2.6.22+ last night and the soft-shutdown (through menu) works with no spinning sound anymore (and the button shutdown still does). However, the system make a loud spinning sound right after I remove my Buffalo external USB drive (even after I unmounted the drive before removal). It also make a smaller spinning sound when I remove flash drive from USB (and the flash drive light doesn't go off after remounting, either - although I don't remember if the light go off or not before my kernel upgrade).

the kernel of Gutsy (2.6.22) solved the problem.

raggar wrote:
> the kernel of Gutsy (2.6.22) solved the problem.
>
> ** Changed in: linux-source-2.6.20 (Ubuntu)
> Status: Confirmed => Fix Released

If 2.6.22 solves the problem, fine.

But why does this lead to the status for linux-source-2.6.20 (Ubuntu)
being changed from "Confirmed" to "Fix Released"? linux-source-2.6.20 is
still broken, isn't it? Shouldn't it be changed to "Won't Fix -- wait
for Gutsy"? Wouldn't that be more honest?

I think that changing the status of the bug is not accurate, feisty users are still affected by this problem and I think that if security bugs get fixed, critical bugs such as this one should be fixed as well.

Gutsy is still on alpha stage and not a lot of people can upgrade the kernel manually, even less people know about the risk this bug implies.

I agree with the last two comments. The fact that this is fixed in the newest kernel is wonderful, but it is not a real fix for users of 2.6.20 on Feisty. Do we really want to recommend to users to add the gutsy sources to sources.list and update their kernel that way?? Some can do that, but we are also the ones who could apply fixes/patches other ways.

The right thing to do, IMHO, is to return this bug to "confirmed" status until a real fix is released until 2.6.20 is patched in our repos, or backport 2.6.22.

Exactly. There are plenty of problems with Gutsy's kernel as I've heard and so it's not wise to think that upgrading to it is a wonderful solution and everyone who experience this bug should do that.

Yes, I know that there are patches one can apply to its system to make the noise go away but:

1) There are quite a few users for whom the patches don't work for some reason or who are just not capable of applying them.
2) Most of the patches don't lose the noise on hibernation/suspend.
3)As said in previous comments: there are many users who aren't even aware of the problem so they can't fix it.

What exactly is the technical difficulty with this bug that prevents the release of a proper patch?

Kaur

Not solved in 2.6.20 apparently.

The problem happens with me too.
Acer Aspire 1692 lmi, 1.73GHzM,Hard Disk hitachi 100GB,1GB ram memory, in my system the partitions appears like SDAx and not like in 6.10(hdax).

Any one have a solution for this?

I'm using a desktop with three IDE hard disks in it, all of which have been detected as /dev/sd* after the switch from Feisty to Gutsy. I can shutdown/hibernate/suspend with no problem. But then again, I've never experienced the clicking sound problem because all the previous kernels I used detected the hard disks as /dev/hd*.

Just to update this bug report - this issue is still apparent. I recently bought a pre-installed Ubuntu Dell laptop Inspiron 6400, and noticed that I had a 'dink' when using shutdown.

A thread from the forums raised the issue (that I wasn't aware of), and following the advice in here, I have applied the temporary fix (it works!, thanks!).

Here is the thread.

http://ubuntuforums.org/showthread.php?t=566072

Nick

Lethe, if you bought a pre-installed Ubuntu laptop, then it's probably using Feisty. The bug seems to be gone in Gutsy, which in prerelease state.

Yes, Dell inspiron 1420 also has a problem.

This bug has been solved with the 2.6.22 kernel in all the distros I've tried that had the problem before. So if you are using any kernel that is older than that (gutsy is the first to carry it in the ubuntu line), then you are still going to experience the problem. AFAIK, no backporting was done on this fix. Meaning, either someone will need to backport it to feisty, or you'll have to upgrade to gutsy as soon as it's released.

Or you can upgrade to the beta now. There are very few issues left in Gutsy. After all, it'll be released in a week and a half from now.

There are instructions to how to upgrade to gutsy's kernel on feisty on the
ubuntuforums site.

On 10/7/07, hyperair <email address hidden> wrote:
>
> Or you can upgrade to the beta now. There are very few issues left in
> Gutsy. After all, it'll be released in a week and a half from now.
>
> --
> bad hard disk noise on shutdown
> https://bugs.launchpad.net/bugs/67810
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>

It's nice to see the kernel change has fixed this issue but as you guys pointed out, it still exists in .20. Fiesty's support is supposed to last for some time and the bug-fixes should be kept active.

Well the workaround which has been shown above works. I've a friend who had confirmed that it worked on his computer. Perhaps sometime soon the backports team will backport the Gutsy kernel over.

when i shutdown my laptop it still have a "PARK" sound
after apply the fix it wont work . too

I'. using Lenovo Y400
HDD is Western D 80GB SATA

this is the fix i apply

Simply write a file at "/etc/rc0.d/S00hdd-shutdown-workaround" which includes this two lines:

#!/bin/sh
echo 1 > /sys/class/scsi_disk/0\:0\:0\:0/stop_on_shutdown

If you have more than one directory at "/sys/class/scsi_disk/" try to find out which of it is your internal harddisk (e.g. "cat /sys/class/scsi_disk/a:b:c:d/device/model") If you have more than one harddisk add another line for each of them.

Then make it executable ("chmod +x filename")

Any Idea this is killing me.......

Perhaps you didn't go throuigh the instructions properly. Could you post the EXACT steps which you followed? By exact steps I mean complete with the values that apply for your computer.

Oh just so you know, it should be "chmod +x /etc/rc0.d/S00hdd-shutdown-workaround". You're supposed to replace "filename" with the appropriate filename.

keatliang2005, I think that a slight click on shutdown is perfectly normal. Still, the sound you are hearing may not be the usual shutdown click but the bug. As far as I know you can pretty easily make sure if you are suffering from the bug or not:

1)Go to terminal, press F11 (to achieve full screen view) and type:

2)sudo apt-get install smartmontools

3) sudo smartctl --device=ata --attributes /dev/sda (This command presumes that your hdd is indeed known as sda)

4)Now look at the raw value of Power-Off_Retract_Count

5) Shut your system down by typing 'sudo shutdown -h now' (yes, 'now' is needed in the command as it sets the time of the shutdown) Then wait until the system shuts down and start it up again

6) Now repeat steps 3-4 and see if the number has grown. If it has, there's definitely something wrong but if it hasn't everything is probably fine with the parking.

Still, I have to say that I am definitely not an expert and so my conclusions may be wrong. It would be nice if someone would confirm my theory or state against it.

erm...... is ok already the Power-Off_Retract_Count wont increase any more but the sound "tick" still present wired....

The tick sound is probably during the proper shutdown of the hard disk. After all, the hard disk must park its read-head before it powers off so that in case of any jarring/jolting, the damages to the hard disk can be minimized.

Power-Off_Retract_Count did not increase, is this mean the HDD park perfectly before power off??

Yes I think it does.

Hope SO。。。。。

maybe it's a result of huge massives of file storing as you do in windows...
i dont believe people suffer from that weird sound if they have an extra partion for large binary objects.

I am assigning this bug to the 'ubuntu-kernel-team' per their bug policy. For future reference you can learn more about their bug policy at https://wiki.ubuntu.com/KernelTeamBugPolicies .

The Hardy Heron Alpha series was recently released which contains an updated version of the kernel. You can download and try the new Hardy Heron Alpha release from http://cdimage.ubuntu.com/releases/hardy/ . You should be able to then test the new kernel via the LiveCD. If you can, please verify if this bug still exists or not and report back your results. General information regarding the release can also be found here: http://www.ubuntu.com/testing/ .

Also note that this report will be kept open against the actively developed kernel but against 2.6.17 and 2.6.20 this will be closed. Thanks.

Just tested the Hardy Heron Alpha6 release on LiveCD, and this bug seems to not appear anymore for me.
The HDD (EIDE) emits the little park sound, about 2 sec before the system powers off.

Hi Nicolas. Thanks for testing and the feedback. Since there was no other comments I'm marking this Fix Released.

Per a decision made by the Ubuntu Kernel Team, bugs will longer be assigned to the ubuntu-kernel-team in Launchpad as part of the bug triage process. The ubuntu-kernel-team is being unassigned from this bug report. Refer to https://wiki.ubuntu.com/KernelTeamBugPolicies for more information. Thanks.