Changes between Initial Version and Version 1 of FAQ


Ignore:
Timestamp:
09/14/2009 10:06:03 AM (10 years ago)
Author:
Gabriele Pohl
Comment:

Transfer content from homepage to here

Legend:

Unmodified
Added
Removed
Modified
  • FAQ

    v1 v1  
     1{{{
     2#!div style="border: 1pt solid; float:right; padding:10px; background-color: #E6E6FA; margin-left:30px;"
     3'''Menu'''[[BR]]
     4[[BR]]
     5[wiki:Download Download Instructions][[BR]]
     6[wiki:TocDoc Documentation][[BR]]
     7[wiki:Help][[BR]]
     8[wiki:TocSupport Device Support][[BR]]
     9[wiki:TocDeveloper Developers Area][[BR]]
     10[wiki:Links]
     11}}}
     12
     13= Smartmontools Frequently Asked Questions (FAQ) =
     14
     15[[PageOutline(2-3,Table of Contents, inline)]]
     16
     17----
     18
     19== Attributes ==
     20
     21----
     22
     23=== I see some strange output from {{{smartctl}}}.  What does it mean? ===
     24
     25The raw SMART attributes (temperature, power-on lifetime, and so
     26on) are stored in vendor-specific structures. Sometime these are
     27strange. Hitachi disks (at least some of them) store power-on
     28lifetime in minutes, rather than hours (see next question below).
     29IBM disks (at least some of them) have three temperatures stored
     30in the raw structure, not just one. And so on.
     31
     32If you find strange output, or unknown attributes, have a look
     33at our wiki pages, were we collect vendor specific info:
     34
     35 * [wiki:AttributesFujitsu Fujitsu]
     36 * [wiki:AttributesIBM IBM (Hitachi)]
     37 * [wiki:AttributesMaxtor Maxtor]
     38 * [wiki:AttributesSeagate Seagate]
     39 * [wiki:AttributesWestern-Digital Western Digital]
     40
     41When you don't find an answer to your question there, please send an email to
     42[https://lists.sourceforge.net/lists/listinfo/smartmontools-support smartmontools-support] and we'll help you try and figure it out.
     43
     44----
     45
     46=== Why is my disk temperature s reported by {{{smartd}}} as 150 Celsius? ===
     47
     48It's not.  Please read the end of the {{{smartd}}} man page (NOTES).
     49For example, in the message:[[BR]]
     50{{{
     51 Device: /dev/hda, SMART Attribute: 194 Temperature_Celsius changed from 94 to 93
     52}}}
     53the value given is the ''Normalized'' not the ''Raw'' Attribute value (the
     54disk temperature in this case is about 22 Celsius).  The
     55{{{'-R'}}} and {{{'-r'}}} Directives modify this behavior, so that
     56the information is printed with the ''Raw'' values as well, for example:
     57{{{
     58 Device: /dev/hda, SMART Attribute: 194 Temperature_Celsius changed from 94 [Raw 22] to 93 [Raw 23]
     59}}}
     60Here the Raw values are the actual disk temperatures in Celsius.  The
     61way in which the Raw values are printed, and the names under which the
     62Attributes are reported, is governed by the various
     63{{{'-v Num,Description'}}} Directives described in the {{{smartd}}}
     64man page. Please see the {{{smartctl}}} manual page for further
     65explanation of the differences between ''Normalized'' and ''Raw'' Attribute values.
     66
     67----
     68
     69=== {{{smartctl}}} reports the age as thousands of hours for my Maxtor/Hitachi/Fujitsu disk , yet it is only a few days old ===
     70
     71On recent disks, Maxtor has started to use Attribute 9 to
     72store the power-on disk lifetime in minutes rather than hours.
     73In this case, use the: {{{'-v 9,minutes'}}} option to correctly
     74display hours and minutes.
     75
     76Some models of Fujitsu disks use Attribute 9 to store
     77the power-on disk lifetime in seconds. In that case, use the:
     78{{{'-v 9,seconds'}}} option to correctly display hours, minutes and seconds.
     79
     80----
     81
     82=== The power-on timer (Attribute 9 raw value) on my Maxtor disk acts strange. ===
     83
     84There are three related problems with Maxtor's SMART firmware:
     85
     86 1. On some Maxtor disks, the raw value of Attribute 9 (Power On Time) is ''supposed'' to be minutes. But it advances at an unpredictable rate, always more slowly than one count per minute. This is because when the disk is in idle mode, the counter stops advancing. This is only supposed to happen in standby mode. This will be corrected in Maxtor product lines released after October 2004.[[BR]][[BR]]
     87 1. In Maxtor disks that use the raw value of Attribute 9 as a minutes counter, only two bytes (of the six available) are used to store the raw value.  So it resets to zero once every 65536=2^16^ minutes, or about once every 1092 hours. This is fixed in all Maxtor disks manufactured after July 2003, where the raw value was extended to four bytes.[[BR]][[BR]]
     88 1. In Maxtor disks that use the raw value of Attribute 9 as a minutes counter, the hour time-stamps in the self-test and ATA error logs are calculated by right shifting 6 bits.  This is equivalent to dividing by 64 rather than by 60.  As a result, the hour time stamps in these logs advance 7% more slowly than they should.  Thus, if you do self-tests once per week at the same time, instead of the time-stamps being 168 hours apart, they are 157 hours apart.  This is also fixed in all Maxtor disks manufactured after July 2003.
     89
     90----
     91
     92
     93== Configuration ==
     94
     95== Protocols, Devices and Controllers ==
     96
     97== Smartmontools Database ==
     98
     99== Selftests ==
     100
     101== Operating System ==
     102
     103== Firmware Issues ==
     104
     105== Distribution ==
     106
     107{{{
     108#!html
     109
     110<!-- Selftests time stamp don't correspond to power-on time -->
     111<h3><a name="time-stamps-WD"></a>The time stamps in the self-test log of my Western Digital (WD) disk
     112don't correspond to the power-on time when the test was run.</h3>
     113
     114<p>
     115The self-test log timestamps in many WD disks roll back to zero every
     1161092 hours (65536 minutes).  This problem is due to a WD firmware bug.
     117The power-on lifetime in hours is correctly stored in Attribute
     1189. However when the power-on lifetime is calculated for self-test log
     119entries, the lifetime in minutes is put into a 16-bit register then
     120divided by 60.  The 16-bit register overflows and wraps around every
     1211092 hours.</p>
     122
     123<p>For WD drives that exhibit this firmware bug, the relationship between
     124Attribute 9's raw value (H) and the time-stamps in the self-test log (h) are given by:<br />
     125Let H = power on hours as shown by Attribute 9 (correct)<br />
     126Let M = 60*H (power on minutes, correct)<br />
     127Let m = M mod 65536 (incorrect value of power on minutes)<br />
     128Let h = m/60 (incorrect value of power on hours, shown in self-test log)</p>
     129<hr />
     130
     131<!-- Normalized values larger than current values -->
     132<h3><a name="worst-larger-than-current"></a>The (normalized) WORST Attribute values of my Western Digital
     133(WD) disk are <strong>larger</strong> than the (normalized) CURRENT Attribute values</h3>
     134<p>Western Digital firmware initializes SMART Attributes 10, 11, and
     135199 after either 120 spin-ups or 8 power-on hours.  Until that time,
     136they have the uninitialized value 253.</p>
     137<hr />
     138
     139<!-- Attributes not recognized -->
     140<h3><a name="attributes-not-recognized"></a>What Attributes does smartmontools not yet recognize?</h3>
     141<p>From Maxtor disks (99), (100), and (101).  These are not used by
     142Maxtor in SMART revision 5.  They will be used in SMART revision 6,
     143but the engineering group has not yet decided what to monitor with these Attributes.</p>
     144<hr />
     145
     146<!-- Operating system requirements -->
     147<h3><a name="os-requirements">What are the operating system requirements?</a></h3>
     148
     149<p>Please see the first section of the
     150<a href="http://sourceforge.net/browser/trunk/smartmontools/INSTALL">
     151INSTALL</a> file.</p>
     152<hr />
     153
     154<!-- manufacturer-specific disk-testing utilities -->
     155<h3><a name="UBCD"></a><b>Where can I find manufacturer-specific
     156disk-testing utilities?</b></h3>
     157
     158<p>A good listing of such utilities can be found <a
     159href="http://www.benchmarkhq.ru/english.html?/be_hdd2.html"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />here</a>.
     160Unfortunately most of these are for MS operating systems, but most can
     161be run from a MS-DOS boot disk.
     162The <a href="http://ubcd.sourceforge.net/"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />UBCD (Ultimate Boot CD)</a>
     163
     164includes most of these disk-testing utilities and many other useful
     165diagnostic tools ready to boot from CD or USB memory stick.
     166The UBCD can be customized by adding other images, like one
     167<a href="faq.html#bootable">containing smartmontools</a>.
     168</p>
     169
     170<p>Note: if you do run one of these utilities, and it identifies the
     171meanings of any SMART Attributes that are not known to smartmontools,
     172please report them to the
     173<a href="https://lists.sourceforge.net/lists/listinfo/smartmontools-support"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />smartmontools-support mailing list</a>.</p>
     174
     175<p>These utilities have an important role to fill.  If your disk has
     176bad sectors (for example, as revealed by running self-tests with
     177smartmontools) and the disk is not able to recover the data from those
     178sectors, then the disk will <em>not</em> automatically reallocate those
     179damaged sectors from its set of spare sectors, because
     180forcing the reallocation to take place may entail some loss of data.
     181Because the commands that force such reallocation are
     182<em>Vendor Specific</em>, most manufactuers provide a utility for this
     183purpose. It may cause data loss but can repair damaged sectors (at
     184least, until it runs out of replacement sectors).</p>
     185
     186<hr />
     187
     188<!-- Can't locate module block-major-65 -->
     189<h3><a name="block-major-65"></a>I found in syslog: '<span class="code">Can't locate module block-major-65</span>'</h3>
     190
     191<p>When I run <tt>smartd</tt>, the SYSLOG <span class="code">/var/log/messages</span>
     192contains messages like this:
     193<br />'<span class="code">smartd: Reading Device /dev/sdv
     194modprobe: modprobe: Can't locate module block-major-65</span>'</p>
     195
     196<p>This is because when <tt>smartd</tt> starts, if there is no
     197configuration file, it looks for all ATA and SCSI devices to monitor
     198(matching the pattern <tt>/dev/hd[a-t]</tt> or
     199<tt>/dev/sd[a-z]</tt>). The log messages appear because your
     200system doesn't have most of these devices.</p>
     201
     202<p>The solution is simple: use the <tt>smartd</tt> configuration file
     203<tt>/etc/smartd.conf</tt> to specify which devices to monitor.</p>
     204
     205<hr />
     206
     207<!-- IBM-Firmware -->
     208<h3><a name="ibm-firmware"></a>What's the story on IBM SMART disks?</h3>
     209
     210<p>Apparently some of the older SMART firmware on IBM disks can
     211interfere with the regular operation of the disk. If you have this
     212problem, here are some links to an IBM Firmware Upgrade that fixes the problem:</p>
     213<ul>
     214    <li><a href="http://www.geocities.com/dtla_update/"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />Geocities Site</a></li>
     215    <li><a href="http://www-3.ibm.com/pc/support/site.wss/document.do?lndocid=MIGR-42215"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />IBM Site #1</a></li>
     216    <li><a href="http://www-1.ibm.com/support/docview.wss?uid=psg1MIGR-42215"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />IBM Site #2</a></li>
     217
     218</ul>
     219<hr />
     220
     221<!-- Check Signature -->
     222<h3><a name="check-signature"></a>How can I check that the package hasn't been tampered with?</h3>
     223
     224<p>Since the <tt>smartmontools</tt> utilities run as root, you might
     225be concerned about something harmful being embedded within
     226them. Starting with release 5.19 of <tt>smartmontools</tt>, the .rpm
     227files and tarball have been GPG signed. The tarball's fingerprint is
     228given in a file on the release page with a name like
     229<tt>smartmontools-5.32.tar.gz.asc</tt>. </p>
     230
     231Please verify these using the
     232<ul>
     233<li><a href="SmartmontoolsSigningKey_2005.txt">Smartmontools GPG Signing Key (current)</a></li>
     234<li><a href="SmartmontoolsSigningKey.txt">Smartmontools GPG Signing Key (before 2005)</a></li>
     235</ul>
     236<hr />
     237
     238<!-- Bootable CD -->
     239<h3><a name="bootable-cd"></a>Is there a bootable standalone CD or floppy that contains smartmontools?</h3>
     240
     241<p>Yes there are. Look to section <a href="http://sourceforge.net/wiki/Download#a4.RunsmartmontoolsfromLive-system">Run from Live-system</a> on the download page.</p>
     242
     243<hr />
     244
     245<h3><a name="testinghelp"></a>Smartmontools for FireWire, USB, and SATA disks/systems</h3>
     246<p>As for USB and FireWire (IEEE 1394) disks and tape drives, the news
     247is not good. They appear to the operating system as SCSI devices but their
     248implementations do not usually support those SCSI commands needed by
     249smartmontools. A consortium associated with IEEE 1394 certified <em>some</em> external enclosures
     250(containing a ATA disk and a protocol bridge) as being compliant to the relevant
     251standards. Even still, that compliance means that they tend to only
     252support the bare minimum of commands
     253needed for device operation (i.e. SMART support is an unsupported
     254extra). Hopefully external USB and Firewire devices will support SAT in
     255the future, see below. Some USB device based on cypress chips support a
     256proprietary protocol (ATACB) that allow to send raw ATA commands (i.e.
     257SMART support).</p>
     258
     259<p>Smartmontools should work correctly with SATA drives under both
     260Linux 2.4 and 2.6 kernels. Depending on which subsystem the SATA
     261controller is in (i.e. <span class="code">drivers/ide</span>,
     262<span  class="code">drivers/ata</span> or libata
     263(under <span  class="code">drivers/scsi</span>) a
     264SATA drive will appear as <span class="code">/dev/hd*</span>
     265
     266or <span  class="code">/dev/sd*</span>. Either way,
     267smartmontools should be able to figure out what is going on and act
     268accordingly. In some cases smartmontools may need a hint in the form of
     269a '<span  class="code">-d sat</span>' or '<span
     270  class="code">-d ata</span>' option on the <tt>smartctl</tt> command line or in the
     271<tt>/etc/smartd.conf</tt> file.
     272There may be a hint to add one of those options in the log file when <tt>smartd</tt> is run as a daemon or on
     273the command line with <tt>smartctl</tt>.
     274The '<span  class="code">-d ata</span>' option means
     275that even though the drive has a SCSI device name, treat it as an ATA
     276disk. Unfortunately such an approach doesn't often work. The next
     277paragraph has more information about '<span  class="code">-d sat</span>'.</p>
     278
     279<p>The SCSI to ATA Translation (SAT) standard (ANSI INCITS 431-2007)
     280may solve many problems in this area. It defines how SCSI commands will
     281be translated to the corresponding ATA commands and defines a
     282pass-through mechanism. ATA commands are conveyed natively by two
     283transports: parallel and serial ATA. SCSI commands can be
     284conveyed by many transports: the veteran SCSI Parallel Interface
     285(SPI), Fibre Channel (FC), Infiniband (SRP), Serial
     286Attached SCSI (SAS), IP (iSCSI and iSER), USB (mass storage), and IEEE
     2871394 (SBP) to name some. Due to their cost and storage capacity, more
     288and more ATA disks (especially SATA disks) are appearing "behind" a
     289SCSI transport. This is especially true of the SAS transport which can
     290painlessly accomodate both SAS and SATA disks. Enter another acronym:
     291SATL which stands for SCSI to ATA Translation Layer. In Linux libata
     292has a SATL in it. Some SAS host bus adapters have a SATL in their
     293firmware. FC might have a SATL in a switch. Perhaps in the future USB
     294and IEEE 1394 enclosures will have a SATL in them. Starting from
     295smartmontools versions 5.36 and 5.37, no matter where a SATL is,
     296irrespective of the operating system in use, the user should have less
     297problems with ATA disks, no matter which transport is involved. As
     298always, it helps to know a little of what is happening under the
     299covers. The '<span class="code">-d sat</span>'
     300option instructs <tt>smartctl</tt>
     301and <tt>smartd</tt> to assume a
     302SATL is in place and act accordingly. The <tt>smartctl</tt> command can often
     303detect a SATL and autoconfigure while in smartmontools version 5.37 <tt>smartd</tt> often needs a hint.</p>
     304
     305<p>The current USB mass storage specification is based on a version of SCSI
     306(SPC-2) that can't support SAT. But some chips manufacturers implement
     307proprietary SCSI commands that allow ATA pass through (similiar like for SAT).
     308Well known is the cypress chipset, that contains an ATACB proprietary pass through
     309(for ATA commands passed through SCSI commands) for which
     310<a href="http://download.cypress.com.edgesuite.net/design_resources/datasheets/contents/cy7c68300c_8.pdf" target="_blank">
     311<img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />some
     312information is publicly available <img src="pics/datei16x16-pdf01.gif" width="16" height="16" border="0" alt="PDF" /></a>.
     313Smartmontools SVN version support
     314these cypress chips via the '<tt>-d usbcypress</tt>' option on the smartctl
     315command line. There is no autodetection at the moment. If you want to know,
     316wether your device supports it, check your device usb id (most
     317cypress usb ata bridge got <tt>vid=0x04b4</tt>, <tt>pid=0x6830</tt>) or to try to call <tt>smartctl</tt>
     318with option '<tt>-d usbcypress</tt>'. If the usb device doesn't support ATACB, smartmontools will abort.</p>
     319
     320<hr />
     321
     322<h3><a name="scsi"></a>Smartmontools for SCSI disks and tapes (TapeAlert)</h3>
     323
     324<p>Smartmontools for SCSI disks and tapes (including medium changers) is
     325discussed on a separate <a href="smartmontools_scsi.html" target="_blank">page</a>.</p>
     326<hr />
     327
     328<!-- Disks behind RAID-Controllers -->
     329<h3><a name="FAQ-RAID"></a>Can I monitor disks behind RAID controllers?</h3>
     330
     331<p>Support for disks behind RAID controllers is highly dependent on both platform
     332and controller type. See our page about <a href="docs/raid-controller_support.html">smartmontools RAID controller support</a>
     333
     334for the details.</p>
     335<hr />
     336
     337<!-- Windows -->
     338<h3><a name="windows"></a>Does it work on Windows?</h3>
     339
     340<p>Yes, finally it does. A windows port of <tt>smartctl</tt> 5.26 by
     341<a href="http://sourceforge.net/users/chrfranke/">Christian Franke</a>
     342was first checked in 2004/02/23 on CVS branch
     343<a href="http://sourceforge.net/browser/branches/RELEASE_5_26_WIN32_BRANCH/sm5">
     344RELEASE_5_26_WIN32_BRANCH</a> and has been merged to the CVS trunk later.</p>
     345
     346<p>The <a href="http://www.cygwin.com/"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />Cygwin</a> environment can be
     347used to built both Cygwin and Windows (using <a href="http://www.mingw.org/"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />MinGW</a>)
     348versions of <tt>smartctl</tt> and <tt>smartd</tt>.
     349Installation instructions for binary distributions can be found
     350<a href="http://sourceforge.net/wiki/Download#InstalltheCygwinpackage-WindowswithCygwininstalled">here for Cygwin</a>
     351and <a href="http://sourceforge.net/wiki/Download#Windows-InstalltheWindowspackage">here for Windows</a>.</p>
     352
     353<hr />
     354
     355<!-- Change of version scheme -->
     356<h3><a name="version-scheme-change"></a>Why did the release version scheme change?</h3>
     357
     358<p>It was non-standard.  So with the move to GNU Autoconf and GNU
     359Automake it changed from 5.X-Y (where X and Y are one or more digits)
     360to 5.Y. Starting with the first release, and moving forward in time, the releases are
     361numbered as follows:</p>
     362<pre>
     3635.0-1,
     3645.0-2,
     365...,
     3665.0-45,
     3675.1-1,
     368...,
     3695.1-18,
     3705.19,
     3715.20,
     372...
     373</pre>
     374<hr />
     375
     376<!-- Add drive to database -->
     377<h3><a name="FAQ-database"></a>My ATA drive is not in the <tt>smartctl</tt>/<tt>smartd</tt> database.  Does this break anything? How do I get it added?</h3>
     378
     379<p> If your drive is not in the database, then the
     380 <i>names</i> of the Attributes (displayed in the <tt>ATTRIBUTE_NAME</tt> column of
     381 <tt>smartctl -A /dev/hd?</tt>) and the <i>format</i> of the the raw Attribute
     382 values shown in the <tt>RAW_VALUE</tt> column may be incorrect.  This
     383 is mostly cosmetic: the essential drive health monitoring/testing
     384 functionality of <b>smartmontools</b> does <i>not</i> depend upon the
     385 database.
     386
     387</p>
     388
     389<p><b>If your drive is not in the database, pleaes check the <a
     390 href="http://sourceforge.net/project/showfiles.php?group_id=64297">sourceforge project page</a>
     391 to be sure that you are using the latest smartmontools release.  Each
     392 new release has additional drives added to the database.  Please do
     393 not submit a new drive for the database without checking to see if it
     394 is already in the database of the current smartmontools release
     395 version.</b></p>
     396
     397<p><b> If your drive is not in the database of the current release,</b>
     398 to have it added to the database, first use the command:<br/>
     399 <span class="code">smartctl -t short /dev/hd?</span><br/> to run a short self-test on
     400 the drive, and wait a few minutes for the test to complete.  Then
     401 email the entire output from:<br/> <span class="code">smartctl -a /dev/hd?</span><br/>
     402
     403 to <a
     404 href="https://lists.sourceforge.net/lists/listinfo/smartmontools-database">smartmontools-database</a>
     405 as a plain-text ASCII email attachment (file type: ".txt").  The timestamp
     406 in the self-test log will help us to determine whether Attribute 9 is
     407 being used to store the lifetime in hours, minutes, or seconds.</p>
     408
     409<p>If you need to use any of the vendor-specific display options
     410 (<tt>-v</tt> options) with the drive, or if any of the Attributes are
     411 behaving strangely, please include that information as well.</p>
     412<hr />
     413
     414<!-- Selftest failed but status is 'PASS' -->
     415<h3><a name="selftest-failed"></a>My ATA drive is failing its self-tests, but its SMART health status is 'PASS'. What's going on?</h3>
     416
     417<p>If your ATA drive supports self-tests, you should run them on a
     418regular basis, for example one per week:
     419<br/><span class="code">smartctl -t long /dev/hd?</span><br/>
     420After the test has completed, you should examine the results with:
     421<br/><span class="code">smartctl -l selftest /dev/hd?</span></p>
     422
     423<p>If the drive fails a self-test, but still has '<tt>PASS</tt>' SMART health
     424status, this usually means that there is a corrupted (uncorrectable=UNC) sector on the
     425disk. This means that the ECC data stored at that sector is not
     426consistent with the user data stored at that sector, and an attempt to read the sector fails with a UNC error.
     427This can be a one-time transient effect: a sudden power failure
     428while the disk was writing to the sector corrupted the
     429ECC code or data, but the sector <em>could</em> correctly store new data.
     430Or it can be a permanent effect: the magnetic media
     431has been damaged by a bit of dust, and the sector could <em>not</em> correctly store new data.</p>
     432
     433<p>If the disk can read the
     434sector of data a single time, and the damage is permanent, not transient, then the disk firmware will mark the
     435sector as 'bad' and allocate a spare sector to replace it.  But
     436if the disk can't read the sector even once, then it won't reallocate
     437the sector, in hopes of being able, at some time in the future, to
     438read the data from it.  <b>A write to an unreadable (corrupted) sector will fix the problem.</b>
     439If the damage is transient, then new consistent data will be written to the sector.
     440If the damange is permanent, then the write will force sector reallocation.
     441Please see <a
     442href="badblockhowto.html" target="_blank">Bad block HOWTO</a>
     443for instructions about how to force this sector to reallocate (Linux only).</p>
     444
     445<p>The disk still has passing health status because the firmware has not
     446found other signs of trouble, such as a failing servo.</p>
     447
     448<p>Such disks can often be repaired by using the disk manufaturer's 'disk
     449evaluation and repair' utility.  Beware: this may force reallocation
     450of the lost sector and thus corrupt or destroy any file system on the
     451disk. See <a
     452href="badblockhowto.html" target="_blank">Bad block HOWTO</a>
     453for generic Linux instructions.</p>
     454
     455<hr />
     456
     457<!-- corrupt sectors -->
     458<h3><a name="corrupt-sectors"></a><tt>smartd</tt> is warning that my ATA disk has unreadable or uncorrectable or pending sectors. What's going on?</h3>
     459
     460<p>
     461Disk drives store data in blocks (sectors) of 512 bytes.  Each 512
     462bytes has additional bytes appended to it (usually 40 to 60) which are
     463used internally by the disk firmware for error checking/detection and
     464correction.  These are called ECC bytes.
     465</p>
     466<p>
     467Sometimes the data in a sector gets corrupted.  This can happen
     468because a speck of dust scratched the disk, or because the disk was
     469powered down while writing data to that sector, or for other reasons.
     470Usually the ECC bytes can be used to correct the corrupted data.
     471However if the ECC bytes are inconsistent or can't be used to correct
     472the bad data, then the 512 bytes of data are lost.  Such a sector is
     473called unreadable or uncorrectable.
     474</p>
     475<p>
     476If your disk has an unreadable sector, this means that some of your
     477data can't be retrieved.  You can force the disk to replace the
     478unreadable sector with a spare good sector, but only at the price of
     479losing the 512 bytes of data forever.
     480</p>
     481
     482<p>
     483Disks with uncorrectable sectors can often be repaired by using the
     484disk manufaturer's 'disk evaluation and repair' utility (see previous
     485FAQ entry).  Beware: this may force reallocation of the lost sector
     486and thus corrupt or destroy any file system on the disk. See <a
     487href="badblockhowto.html">Bad block HOWTO</a>
     488for generic Linux instructions.
     489</p>
     490<p>
     491Normally when an uncorrectable sector is found, the disk puts this
     492onto a 'pending sector list' to indicate that it should be replaced
     493with a spare good sector.  However this replacement won't take place
     494until either the disk can read the data on the bad sector, or is
     495commanded to write new data to that bad sector.
     496</p>
     497<hr />
     498
     499<h3><a name="bios-setting"></a>My computer's BIOS has a SMART enable/disable setting.  What
     500does it do, and how should I set it?</h3>
     501
     502<p>Some type of BIOS can check the SMART health status of a disk at
     503bootup: the equivalent of '<span class="code">smartctl -H /dev/hd?</span>'.  This one-time check on
     504bootup is done if the BIOS SMART setting is set to '<tt>ENABLE</tt>', and is
     505not done if the setting is set to '<tt>DISABLE</tt>'.</p>
     506
     507<p>If this one-time check is done, and the disk's health status is found
     508to be '<tt>FAIL</tt>', then typically the BIOS will display an error message
     509and refuse to boot the machine.</p>
     510
     511<p>For the proper functioning of smartmontools, either BIOS setting may
     512be used.</p>
     513<hr />
     514
     515<h3><a name="fedora-failed"></a>My Fedora Core Linux system displays the startup message: smartd [FAILED]</h3>
     516
     517<p>Fedora Core is distributed with a <tt>smartd</tt> configuration file
     518
     519<tt>/etc/smartd.conf</tt> that monitors the first IDE disk /dev/hda.  If this
     520device does not exist (or lacks SMART capability) you will get the
     521error message above.  Look in SYSLOG (/var/log/messages) for
     522additional details about what is going wrong.</p>
     523
     524<p>The solution: If your system has only SCSI disks, or has IDE disk(s)
     525on a non-primary controller, just edit <tt>/etc/smartd.conf</tt> to reflect the
     526correct location of the drive(s).  Please also read the <tt>smartd.conf</tt>
     527man page for additional information.</p>
     528<hr />
     529
     530<h3><a name="temp-seagate"></a>Attribute 194 (Temperature Celsius) behaves strangely on my Seagate disk</h3>
     531
     532<p>Some Seagate disks store the current temperature Celsius in both the
     533RAW and NORMALIZED Attribute 194 values, and the maximum lifetime
     534temperature in Celsius in the WORST value.  Since cooler is better,
     535this means that in this case, <em>lower</em> NORMALIZED Attribute values
     536are farther from failure, and that over time the WORST Attribute
     537values get <em>larger</em>, not <em>smaller</em> (as with other
     538Attributes).</p>
     539<hr />
     540
     541<h3><a name="ata-error-count"></a>What's this <tt>smartctl</tt> message mean?: Warning: ATA error count 9 inconsistent with error log pointer 5</h3>
     542
     543<p>The ATA error log is stored in a circular buffer, and the ATA
     544specifications are unambiguous about how the entries should be
     545ordered.  This warning message means that the disk's firmware does not
     546strictly obey the ATA specification regarding the ordering of the
     547error log entries in the circular buffer.  Smartmontools will correct
     548for this oversight, so this warning message can be safely ignored by
     549users.  (On the other hand, firmware engineers: please read the ATA
     550specs more closely then fix your code!).</p>
     551<hr />
     552
     553<h3><a name="FAQ-win-ata-as-scsi"></a>On Windows, <tt>smartctl</tt> aborts
     554with the message "...SMART_GET_VERSION failed". What is going wrong?</h3>
     555
     556<p>A failing
     557<a href="http://msdn2.microsoft.com/en-us/library/ms803644.aspx"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />SMART_GET_VERSION</a>
     558call means that the device driver does not
     559implement the I/O controls (see <a href="faq.html#FAQ-win-ioctl">below</a>)
     560to access ATA SMART functionality.</p>
     561
     562<p>Some Windows drivers for (S)ATA controllers are implemented as SCSI
     563class drivers. This is usually the case for drivers which support RAID.
     564Unfortunately, such drivers do not support the ATA specific SMART I/O
     565controls.</p>
     566<hr />
     567
     568<h3><a name="FAQ-win-ioctl"></a>On Windows, <tt>smartctl</tt> prints the
     569message: "...Log Read failed: Function not implemented". What is going wrong?</h3>
     570
     571<p>This means that the device driver does not support the command SMART READ LOG.
     572<span class="marked">The message does not indicate a hard disk problem!</span>
     573It does also not mean that the disk itself does not support SMART logs.
     574It may still be possible to read the logs with a Linux version of smartmontools run from
     575some <a href="faq.html#bootable">bootable CD</a>.</p>
     576
     577<p>To access ATA SMART functionality on Windows, smartmontools uses the
     578I/O control calls
     579<a href="http://msdn2.microsoft.com/en-us/library/ms804269.aspx"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />SMART_RCV_DRIVE_DATA</a> and
     580<a href="http://msdn2.microsoft.com/en-us/library/ms803645.aspx"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />SMART_SEND_DRIVE_CMD</a>.
     581These calls were available since Win95 OSR2.
     582An example program from Microsoft can be found
     583<a href="http://download.microsoft.com/download/winddk/sample3/9x/W9X/EN-US/SmartApp.exe"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />here</a>
     584(the related KB article 208048 is no longer available).</p>
     585
     586<p>Starting with NT4, these calls do more restrictive parameter checks.
     587In particular, the command codes for SMART READ LOG and ABORT SELF-TEST
     588are not accepted. To perform these functions, smartmontools uses the
     589undocumented functions SCSIOP_ATA_PASSTHROUGH (NT4) or
     590IOCTL_IDE_PASS_THROUGH (2000/XP) instead.
     591An example program using these calls can be found
     592<a href="ftp://ftp.heise.de/pub/ct/listings/0207-218.zip"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />here</a>,
     593a related newsgroup thread is
     594<a href="http://groups.google.com/group/microsoft.public.development.device.drivers/browse_frm/thread/e9763262823e11d1"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />here</a>.</p>
     595
     596<p>Unfortunately, these undocumented functions are not implemented in
     597most vendor specific ATA device drivers. <tt>Smartctl</tt> prints a
     598"Function not implemented" message in this case.</p>
     599
     600<p>A new I/O control call
     601<a href="http://msdn2.microsoft.com/en-us/library/ms804551.aspx"><img src="pics/linkext.gif" width="11" height="11" border="0" alt="extLink: " />IOCTL_ATA_PASS_THROUGH</a>
     602is available since Win2003 and XP SP2.
     603It should be supported by most new drivers. Experimental code using
     604this call was added 2006-04-27 and is included in smartmontools
     605release 5.37.</p>
     606
     607}}}