Bug#664041: acpi-support: load cycles problem revisited

Package: acpi-support
Version: 0.138-10
Severity: important

Hi,

Just did a fresh install of Debian Wheezy and noticed that the count
of load cycles of my hard-drive are increasing constantly and at an
undesired frequency.

This issue is famous and was already reported five years ago as #448673

At that time, acpi-support upstream was updated to include some
90-hdparm.sh scripts at /etc/acpi to perform hdparm operations at due
time, which solved the problem for debian package too.

But, Debian´s acpi-support upstream is linked to Ubuntu´s package (as
seen in http://packages.qa.debian.org/a/acpi-support.html). Some time
later, Ubuntu´s own hdparm and pm-utils packages evolved to include a
chain of other scritps that already took care of calling hdparm with
correct parameters, starting from pm-powersave at /etc/acpi/. Those
changes only affected Ubuntu.

So 90hdparm.sh scripts became redundand at Ubuntu, and they were
removed from its package. As it is the upstream for Debian
acpi-support, they were also removed after an update - see Debian
acpi-support 0.129-1 changelog for details.

However, Debian do not have the same solution using pm-utils, so
basically the load cycle bug was free again.

As discussed extensively in the past, this may not even be considered
a bug, and definitely is a hardware manufacturer´s problem. Also, now
there are easy solutions: (i) just installing laptop-mode-tools or
(ii) copying one of the 90-hdparms.sh scripts, now supplied at
/usr/share/doc/acpi-support/examples/acpi/, to /etc/pm/power.d/.

But the fact is, for a fresh new install, even including laptop
packages, some hard drives may start cycling to death. I only noticed
this because of the head parking noise, and when checked already have
~3K cycles in 120 hours of use. Have I noticed only one year later,
maybe the counter was already at number much closer to it´s factory
"limits".

Maybe I'm missing something here, but that's what I've found these days.
Info:

# hdparm -i /dev/sda

/dev/sda:

Model=WDC WD7500BPVT-75HXZT3, FwRev=03.01A03, SerialNo=WD-WX81E81TDZ47
Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=50
BuffType=unknown, BuffSize=8192kB, MaxMultSect=16, MultSect=16
CurCHS=65535/1/63, CurSects=4128705, LBA=yes, LBAsects=1465149168
IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 udma2 udma3 udma4 udma5 *udma6
AdvancedPM=yes: unknown setting WriteCache=enabled
Drive conforms to: Unspecified: ATA/ATAPI-1,2,3,4,5,6,7

* signifies the current active mode


$ uname -a
Linux trouble 3.2.0-1-amd64 #1 SMP Fri Feb 17 05:17:36 UTC 2012 x86_64 GNU/Linux


Many thanks

Igor


<pre>Package: acpi-support
Version: 0.138-10
Severity: important<br><br>Hi,<br><br>Just did a fresh install of Debian Wheezy and noticed that the count of load cycles of my hard-drive are increasing constantly and at an undesired frequency. <br><br>This issue is famous and was already reported five years ago as #448673<br>
<br>At that time, acpi-support upstream was updated to include some 90-hdparm.sh scripts at /etc/acpi to perform hdparm operations at due time, which solved the problem for debian package too.<br><br>But, Debian´s acpi-support upstream is linked to Ubuntu´s package (as seen in <a href="http://packages.qa.debian.org/a/acp...html" target="_blank">http://packages.qa.debian.org/a/acp...</a>). Some time later, Ubuntu´s own hdparm and pm-utils packages evolved to include a chain of other scritps that already took care of calling hdparm with correct parameters, starting from pm-powersave at /etc/acpi/. Those changes only affected Ubuntu.<br>
<br>So 90hdparm.sh scripts became redundand at Ubuntu, and they were removed from its package. As it is the upstream for Debian acpi-support, they were also removed after an update - see Debian acpi-support 0.129-1 changelog for details.<br>
<br>However, Debian do not have the same solution using pm-utils, so basically the load cycle bug was free again.<br><br>As discussed extensively in the past, this may not even be considered a bug, and definitely is a hardware manufacturer´s problem. Also, now there are easy solutions: (i) just installing laptop-mode-tools or (ii) copying one of the 90-hdparms.sh scripts, now supplied at /usr/share/doc/acpi-support/examples/acpi/, to /etc/pm/power.d/.<br>
<br>But the fact is, for a fresh new install, even including laptop packages, some hard drives may start cycling to death. I only noticed this because of the head parking noise, and when checked already have ~3K cycles in 120 hours of use. Have I noticed only one year later, maybe the counter was already at number much closer to it´s factory &quot;limits&quot;.<br>

Maybe I&#39;m missing something here, but that&#39;s what I&#39;ve found these days.
<br>Info:<br><br># hdparm -i /dev/sda<br><br>/dev/sda:<br><br> Model=WDC WD7500BPVT-75HXZT3, FwRev=03.01A03, SerialNo=WD-WX81E81TDZ47<br> Config={ HardSect NotMFM HdSw&gt;15uSec SpinMotCtl Fixed DTR&gt;5Mbs FmtGapReq }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=50<br> BuffType=unknown, BuffSize=8192kB, MaxMultSect=16, MultSect=16<br> CurCHS=65535/1/63, CurSects=4128705, LBA=yes, LBAsects=1465149168<br> IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio3 pio4 <br> DMA modes: mdma0 mdma1 mdma2 <br> UDMA modes: udma0 udma1 udma2 udma3 udma4 udma5 *udma6 <br> AdvancedPM=yes: unknown setting WriteCache=enabled<br> Drive conforms to: Unspecified: ATA/ATAPI-1,2,3,4,5,6,7<br>
<br> * signifies the current active mode<br><br><br>$ uname -a<br>Linux trouble 3.2.0-1-amd64 #1 SMP Fri Feb 17 05:17:36 UTC 2012 x86_64 GNU/Linux<br><br><br>Many thanks<span><font color="#888888"><br><br>Igor<br> </font></span></pre>






To UNSUBSCRIBE, email to debian-bugs-dist-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org