Smart error log version 1

under construction, only slightly usable so far, please report errors in the talk page

Disclaimer: this page is based on personal experience gained from examining numerous SMART reports, therefore it should not be considered authoritative. Accuracy however is highly desired, so please feel free to correct it as needed, or suggest corrections or question its statements on the associated Talk page.

Prologue

There is a lot of ignorance and misinformation out there about SMART reports, so this will be an effort to help users to a better understanding of the content of SMART reports.

Consider the following SMART report extract:

Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   112   099   006    Pre-fail  Always       -       42208416
  3 Spin_Up_Time            0x0003   096   096   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       7
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   056   055   030    Pre-fail  Always       -       25772440425
  9 Power_On_Hours          0x0032   100   100   000    Old_age   Always       -       72
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       7
183 Runtime_Bad_Block       0x0032   100   100   000    Old_age   Always       -       0
184 End-to-End_Error        0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
188 Command_Timeout         0x0032   100   100   000    Old_age   Always       -       0
189 High_Fly_Writes         0x003a   100   100   000    Old_age   Always       -       0
190 Airflow_Temperature_Cel 0x0022   057   048   045    Old_age   Always       -       43 (Min/Max 36/43)
191 G-Sense_Error_Rate      0x0032   100   100   000    Old_age   Always       -       0
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       5
193 Load_Cycle_Count        0x0032   100   100   000    Old_age   Always       -       19
194 Temperature_Celsius     0x0022   043   052   000    Old_age   Always       -       43 (0 28 0 0)
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       260348032581703
241 Total_LBAs_Written      0x0000   100   253   000    Old_age   Offline      -       423266408125
242 Total_LBAs_Read         0x0000   100   253   000    Old_age   Offline      -       97907054046

Looks rather intimidating, doesn’t it, with huge scary numbers! But with a little knowledge from this page, you should be able to quickly say «That drive looks fine! A little warm though!»

Introduction to SMART

From SMART on Wikipedia, «S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology; often written as SMART) is a monitoring system for computer hard disk drives to detect and report on various indicators of reliability, in the hope of anticipating failures.» It was a laudable effort by the drive manufacturers to provide standard ways to both report current drive parameters and status, and also indicate issues, especially those that might be predictive of imminent drive failure. Unfortunately, the standard had considerable ambiguity, and the various drive engineers have often differed greatly in their interpretations and implementations of both the common attributes, and the introduction of new attributes.

This page is primarily a guide to understanding SMART attributes, in real world usage. They are unfortunately very inconsistent in their behavior, not only between the different attributes, but between the various drive models, and especially between brands. In some cases, the RAW_VALUE is the counter to watch, in others, it is more important to watch what the VALUE does, and there are yet other behaviors too. To understand a particular attribute report line, you have to understand how that SMART attribute is usually handled, keeping in mind who the manufacturer is, and to a lesser extent, what drive model it is. You can try researching it online, but information is really skimpy, nothing authoritative at all from the manufacturers themselves. The table of SMART attributes below should help you understand them, but every manufacturer uses a different set of SMART attributes, even using the common ones in differing ways, even across their own drive models.

There are many computer professionals with a very low opinion of SMART reporting, and they generally discount SMART reports, partly because of all the inconsistency, but also because many drives fail with no SMART warnings at all. I find that once you understand the inconsistencies, and keep some perspective, there is much that can still be learned. For one example, the Seek_Error_Rate (a critical attribute) on Seagate drives generally starts and stays in the mid 50’s to high 60’s (attribute values generally start at 100 and drop to 1). Not knowing this, you might immediately think there is a serious issue with your new Seagate drive. But now that you do know this, you won’t be concerned until it drops into the low 50’s or below. The same Seek_Error_Rate value on any other brand would be immediately concerning. Hopefully the table below will help you understand what ‘normal’ looks like, for the different attributes on different drives by different makers.

SMART report structure

Each section below includes an example of that section, in a gray box with dotted border. It’s just an example, yours may greatly differ.

General information section

• Identifying information for the SMART program and the drive — its model, serial number, firmware, capacity/size, time of this report, and SMART support status

smartctl 5.39.1 2010-01-28 r3054 [i486-slackware-linux-gnu] (local build)
Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net

=== START OF INFORMATION SECTION ===
Device Model:     ST1500DL003-9VT16L
Serial Number:    5YD3D71H
Firmware Version: CC32
User Capacity:    1,500,301,910,016 bytes
Device is:        Not in smartctl database [for details use: -P showall]
ATA Version is:   8
ATA Standard is:  ATA-8-ACS revision 4
Local Time is:    Fri Nov 18 16:11:43 2011 EST
SMART support is: Available - device has SMART capability.
SMART support is: Enabled 

SMART overall health test

• Basic overall health test of the drive, only 2 choices — PASSED or FAILED
• If test result is FAILED, then that means the SMART firmware believes that the drive is in imminent danger of catastrophic failure, so it is imperative to copy off ALL important data. Usually, it is best to copy off the most important files, then the next most important files, then the next, and so on, because the drive may completely quit before you finish copying.

SMART overall-health self-assessment test result: PASSED 

SMART parameters section

• These are generally of little interest to us
• They do include the recommended polling time for the short and long tests, in other words don’t request a SMART report any sooner than this recommendation
  • Unfortunately the original standard must have stipulated using a single byte to store the polling times, which caps their maximum value at 255. That makes the ‘Extended self-test’ (the long test) polling time of 255 rather useless.
• I have seen a case where an unusually long ‘Total time to complete Offline data collection’ for one unusually slow drive was the only indication of a faulty drive. The SMART reports for other drives that were exactly the same model had essentially identical SMART reports, with no issues, except for the difference in this parameter.

Offline data collection status:  (0x82)	Offline data collection activity
					was completed without error.
					Auto Offline Data Collection: Enabled.
Self-test execution status:      (   0)	The previous self-test routine completed
					without error or no self-test has ever 
					been run.
Total time to complete Offline 
data collection: 		 ( 623) seconds.
Offline data collection
capabilities: 			 (0x7b) SMART execute Offline immediate.
					Auto Offline data collection on/off support.
					Suspend Offline collection upon new
					command.
					Offline surface scan supported.
					Self-test supported.
					Conveyance Self-test supported.
					Selective Self-test supported.
SMART capabilities:            (0x0003)	Saves SMART data before entering
					power-saving mode.
					Supports SMART auto save timer.
Error logging capability:        (0x01)	Error logging supported.
					General Purpose Logging supported.
Short self-test routine 
recommended polling time: 	 (   1) minutes.
Extended self-test routine
recommended polling time: 	 ( 255) minutes.
Conveyance self-test routine
recommended polling time: 	 (   2) minutes. 

SMART attributes section

• This is the table of SMART attributes for this drive. The columns are described below the example. Yours may greatly differ from this example, as some drives report more attributes, and some drives report considerably fewer. The newest drives often introduce new attributes.

SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   111   100   006    Pre-fail  Always       -       32796080
  3 Spin_Up_Time            0x0003   095   095   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       5
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   100   253   030    Pre-fail  Always       -       265367
  9 Power_On_Hours          0x0032   100   100   000    Old_age   Always       -       19
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       5
183 Runtime_Bad_Block       0x0032   100   100   000    Old_age   Always       -       0
184 End-to-End_Error        0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
188 Command_Timeout         0x0032   100   100   000    Old_age   Always       -       0
189 High_Fly_Writes         0x003a   100   100   000    Old_age   Always       -       0
190 Airflow_Temperature_Cel 0x0022   070   069   045    Old_age   Always       -       30 (Lifetime Min/Max 26/31)
191 G-Sense_Error_Rate      0x0032   100   100   000    Old_age   Always       -       0
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       4
193 Load_Cycle_Count        0x0032   100   100   000    Old_age   Always       -       5
194 Temperature_Celsius     0x0022   030   040   000    Old_age   Always       -       30 (0 26 0 0)
195 Hardware_ECC_Recovered  0x001a   037   029   000    Old_age   Always       -       32796080
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       172868138696723
241 Total_LBAs_Written      0x0000   100   253   000    Old_age   Offline      -       2919100768
242 Total_LBAs_Read         0x0000   100   253   000    Old_age   Offline      -       572998840 

• Column 1 is the attribute number, usually a decimal number between 1 and 255. Some SMART tools report it in hex, from 01 to FF. These are relatively standard ID’s, except that different manufacturers will occasionally introduce a new one, unused by anyone else. Generally, the only ones you can count on seeing are: 1, 3, 4, 5, 7, 9, 10, 187, 190 or 194, 193, 195, 197, 198, and 199.
• Column 2 is the relatively standardized attribute name. There are a few that seem only used by a single manufacturer.
• Column 3 is the attribute handling flag, of no interest to us — ignore it.
• Column 4 is the VALUE, one of the most important values in the table. It is stored in a single byte on the drive for each SMART attribute, so its range is from 0 to 255.
  • However, the values of 0, 254, and 255 are reserved for internal use, so you never see them.
  • The value of 253 usually always means «Not Used Yet», so when you see it, you are probably looking at a brand new drive. Sometimes though, there can be a few attributes that take awhile before they are used, so may stay 253 for longer.
  • VALUE is almost always used as a normalized scale of perfectly good to perfectly bad, usually starting at VALUE=100, then dropping toward a worst case of VALUE=1. You can generally think of it as representing a scale starting at 100% good, then slowly dropping until failure at some predetermined percentage number, in the THRESHOLD column.
  • Someone realized that if the values only run from 100 to 1, then they are wasting the possible values from 101 to 252, so some SMART programmers have decided to stretch the scale for certain attributes to start at 200 instead of 100, providing twice the data points. Unfortunately, which attributes are scaled from 200 to 1 is completely inconsistent, with almost all SMART reports showing some attributes starting at 100, and other attributes starting at 200. In addition, there are a few Maxtor and Samsung drives that took the start of the scale all the way to 252 or 253! Above, you see all but 1 attribute using 100, the exception being attribute 199 which starts at 200. In general, you can think of 200-type scales as 100 times 2 (just divide the number by 2), and from now on, that is what we are going to do in most of the discussion.
  • The temperature attributes 190 and 194 are exceptions to the scaling. They are either temperatures or forms of the temperature, and they don’t scale (their WORST value may look like it scales though).
  • The error rate attributes 1 and 7 are also exceptions, although of a different kind. Raw read and seek errors are a natural part of normal operation, so even in a brand new and perfect drive, there is a factory-determined optimal rate of read and seek errors. They are nothing to worry about, they’re the natural result of temperature expansion and other things, and they are used to help the drive constantly recalibrate itself. But because these error rates are non-zero, you essentially cannot have a perfect error rate of zero that you declare is a VALUE of 100. So manufacturers determine what an optimal error rate should be and call it 100. But often, drives may achieve an error rate (especially when they are new) that is even better than the optimal one set by the manufacturer, which results in an error rate that is HIGHER than 100! For an example, see the VALUE above of attribute 1, the Raw_Read_Error_Rate. It’s as if the drive is performing at 111%!
• Column 5 is WORST, the lowest VALUE ever recorded (except for a few unusual and uncommon cases).
  • [incomplete]
• Column 6 is THRESH, the manufacturer determined lowest value that WORST should be allowed to fall to, before reporting it as a FAILED quantity. Some are counters, some are informational such as temperature or hours used or
  • [incomplete]
• Column 7 is TYPE, the type of attribute. It can either be Pre-fail or Old_age.
  • If it is Pre-fail, then the attribute is considered a critical attribute, one that participates in the overall SMART health assessment (PASSED/FAILED) of the drive. If the value of WORST falls below THRESH, then the drive FAILS the overall SMART health test, and complete failure may be imminent. The Pre-fail term means that if this attribute fails, then the drive is considered ‘about to fail’.
  • If it is Old_age, then the attribute is considered (for SMART purposes) a noncritical attribute, one that does not fail the drive. The Old_age term means that the attribute is related to normal aging, normal wear and tear of the drive.
  • When new attributes are introduced, they may seem like a critical item, perhaps even with an appropriate THRESH set. But if they are marked as Old_age, then they do NOT fail the drive, even if WORST falls below THRESH. Naturally, this could be highly concerning, but there is no authoritative interpretation available, so no definitive conclusions can be made. These attributes should be considered Experimental.
  • [incomplete]
• Column 8 is UPDATED. Supposedly, this is an indicator when the attribute is updated, Always or Offline. If Always, then it is assumed that the attribute is updated whenever a relevant event occurs. In other words, it is always ‘live’. If Offline, then supposedly the attribute is only updated when offline tests are being performed. But in real life, our experience is that these are inaccurate. Just look at the example above, at attributes 241 and 242. They appear to be live counters of LBA’s read and written, yet the test section of that particular SMART report indicates that there have been no offline tests performed!
• Column 9 is WHEN_FAILED, usually and thankfully blank! If not blank, then it indicates the last operational hour (from attribute 9 Power_On_Hours) that this attribute failed!
• Column 10 is RAW_VALUE, a manufacturer controlled raw number, which may or may not be of interest to us. From now on, we will often shorten its name and refer to it only as ‘the RAW’.
  • [incomplete]

Error Log section

• [incomplete]

SMART Error Log Version: 1
No Errors Logged

• [incomplete, need example with errors]

Test results section

• [incomplete]

SMART Self-test log structure revision number 1
No self-tests have been logged.  [To run self-tests, use: smartctl -t]

SMART Selective self-test log data structure revision number 1
 SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
    1        0        0  Not_testing
    2        0        0  Not_testing
    3        0        0  Not_testing
    4        0        0  Not_testing
    5        0        0  Not_testing
Selective self-test flags (0x0):
  After scanning selected spans, do NOT read-scan remainder of disk.
If Selective self-test is pending on power-up, resume after 0 minute delay.

• [incomplete, need example with tests]

Table of attributes

For a fuller description of each attribute, please see Known ATA S.M.A.R.T. attributes on Wikipedia.
[incomplete]

1 Raw_Read_Error_Rate

• This is an indicator of the current rate of errors of the low level physical sector read operations. In normal operation, there are ALWAYS a small number of errors when attempting to read sectors, but as long as the number remains small, there is NO issue with the drive. Error correction information and retry mechanisms are in place to catch and fix these errors. Manufacturers therefore determine an optimal level of errors for each drive model, and set up an appropriate scale for monitoring the current error rate. For example, if 3 errors per 1000 read operations seems near perfect to the manufacturer, then an error rate of 3 per 1000 ops might be set to an attribute VALUE of 100. If the rate increased to 10 per 1000, then the rate might be scaled to 80 (completely under manufacturer control, and NEVER revealed or explained to us!).
• They are called Raw Reads to distinguish them from the more common term ‘read errors’, which represent a much higher level read operation. What we usually refer to as a ‘read error’ is an error returned by a read process, that has attempted a series of one or more seeks and raw reads, plus optional error corrections and retries. It either returns an indicator of total success plus the sector data (considered to be in perfect shape), or it returns an error code, and no sector data.
• PLEASE completely ignore the RAW_VALUE number! Only Seagates report the raw value, which yes, does appear to be the number of raw read errors, but should be ignored, completely. All other drives have raw read errors too, but do not report them, leaving this value as zero only. To repeat, Seagates are not worse than other drives because they appear to have raw read errors, rather they are the only one to report the number. I suspect that others do not report the number to avoid a lot of confusion, and questions for their tech support people. Seagate leaves those of us who provide tech support the job of answering the constant questions about this number. Hopefully now that you understand this, you will never bother a kind IT person with questions about the Raw_Read_Error_Rate RAW_VALUE again?
• [incomplete?]
• Critical attribute — if its WORST falls below its THRESH, then the drive will be considered FAILED

3 Spin_Up_Time

• [incomplete]

4 Start_Stop_Count

• [incomplete]

5 Reallocated_Sector_Ct

• [incomplete]

7 Seek_Error_Rate

• [incomplete]

9 Power_On_Hours

• [incomplete]

[the most important part of this whole page is completely incomplete!]

Additional info

• also known more accurately as S.M.A.R.T. or Self-Monitoring, Analysis and Reporting Technology
• Reference materials
  • http://en.wikipedia.org/wiki/S.M.A.R.T. — all about S.M.A.R.T., from Wikipedia; recommended reading!
  • http://en.wikipedia.org/wiki/S.M.A.R.T#Known_ATA_S.M.A.R.T._attributes — table of S.M.A.R.T. attributes, from Wikipedia
  • http://www.linuxjournal.com/article/6983 — an excellent article on SMART and smartctl, from Linux Journal
  • http://smartmontools.sourceforge.net/ — smartmontools Home Page
  • http://smartmontools.sourceforge.net/faq.html — smartmontools FAQ Page
  • http://smartmontools.sourceforge.net/man/smartctl.8.html — MAN Page for smartmontools
• UnRAID related (some are marked << old >>, meaning some part may be obsolete or incompatible with current releases of UnRAID)
  • http://lime-technology.com/forum/index.php?topic=13054.msg53337#msg53337 — keeping SMART values in perspective, and how to properly interpret them — a series of posts to help users alarmed by the very large numbers they find in a SMART report or ‘diff’
  • http://lime-technology.com/forum/index.php?topic=2135.msg15733#msg15733 — a script for grabbing dated SMART reports for all drives
  • << old >> FAQ#How_can_I_find_out_more_information_about_a_hard_drive.3F — intro to obtaining the SMART info for a drive
  • << old >> FAQ#Why_is_a_temp_not_showing_for_a_drive.3F — enabling SMART so temps can be accessed and displayed
  • Troubleshooting#Hard_drive_failures — has a section on smartctl commands for getting SMART reports, and running tests
  • UnRAID_Add_Ons#UnMENU — the Disk Management plugin has buttons for SMART reports and tests
  • http://lime-technology.com/forum/index.php?topic=2708 — the MyMain thread; an UnMENU plugin; after installing UnMENU, install this next; has a Smart View that provides color-coded SMART info for all drives
  • SmartHistory — a tool for monitoring the SMART parameters of your drives, and provide reporting and notification of changes in SMART attributes; produces customizable reports, with graphing capabilities

[incomplete]

Welcome to the Slackware Documentation Project

Table of Contents

SMART/S.M.A.R.T stands for Self-Monitoring, Analysis and Reporting Technology. It is basically a system that collects information about a hard disk drive (HDD) and solid state drive (SDD), and allows you to run some tests on the drive to determine its approximate health.

It is important to note that SMART is far from perfect. Although a failed “Pre-fail” SMART attribute predicts failure, having no failed attributes does NOT mean the drive is not failing. The drive can be failing with above threshold attributes. This leads us to the next section backing up your data.

According to CERT you should follow the 3-2-1 rule:

3 - Keep 3 copies of any important file: 1 primary and 2 backups.
2 - Keep files on 2 different media types to protect against different types of hazards.
1 - Store 1 copy offsite (e.g. outside your or business facility).

In summary, keep 3 backups: 1 primary, 1 onsite, 1 offsite. This is of critical importance because your device can fail at any time without warning and for various reasons. Backing up your data is the only way to be reasonably sure that you won’t lose it. You CANNOT rely on SMART to reliably tell you when your HDD is going to fail and to do so in due time to allow you to save your data.

In order to be able to use SMART you need:

1. A HDD or SSD that supports SMART

2. SMART enabled in the UEFI/BIOS

3. Software to interface with SMART

Some commonly used software to interface with SMART is smartmontools, or you can find individual manufacturer’s utilities on UBCD. Some people prefer smartmontools because it is easily accessible from the command line. Others prefer the manufacturer’s utilities because they sometimes have more features than smartmontools. Which is better is mostly down to user preference and the details of the situation. For this article we will focus on smartmontools and more specifically smartctl.

In order to display the SMART attributes with smartmontools you need to run the following as root:

smartctl -a /dev/sda

Note that we will be assuming that /dev/sda is your HDD/SSD device node. In many cases this is the first HDD/SSD on the system, but you need to double check to make sure it is the HDD/SSD you are interested in.

The output will be something like:

bash-4.2# smartctl -a /dev/sda
smartctl 5.43 2012-06-30 r3573 [x86_64-linux-3.10.63] (local build)
Copyright (C) 2002-12 by Bruce Allen, http://smartmontools.sourceforge.net

=== START OF INFORMATION SECTION ===
Model Family:     Seagate Barracuda (SATA 3Gb/s, 4K Sectors)
Device Model:     ST1000DM003-1CH162
Serial Number:    Z1D6DR9C
LU WWN Device Id: 5 000c50 064a62447
Firmware Version: CC49
User Capacity:    1,000,204,886,016 bytes [1.00 TB]
Sector Sizes:     512 bytes logical, 4096 bytes physical
Device is:        In smartctl database [for details use: -P show]
ATA Version is:   8
ATA Standard is:  ACS-2 (unknown minor revision code: 0x001f)
Local Time is:    Sun Jan  4 16:02:08 2015 CST
SMART support is: Available - device has SMART capability.
SMART support is: Enabled

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

General SMART Values:
Offline data collection status:  (0x82)	Offline data collection activity
					was completed without error.
					Auto Offline Data Collection: Enabled.
Self-test execution status:      (   0)	The previous self-test routine completed
					without error or no self-test has ever 
					been run.
Total time to complete Offline 
data collection: 		(  584) seconds.
Offline data collection
capabilities: 			 (0x7b) SMART execute Offline immediate.
					Auto Offline data collection on/off support.
					Suspend Offline collection upon new
					command.
					Offline surface scan supported.
					Self-test supported.
					Conveyance Self-test supported.
					Selective Self-test supported.
SMART capabilities:            (0x0003)	Saves SMART data before entering
					power-saving mode.
					Supports SMART auto save timer.
Error logging capability:        (0x01)	Error logging supported.
					General Purpose Logging supported.
Short self-test routine 
recommended polling time: 	 (   1) minutes.
Extended self-test routine
recommended polling time: 	 ( 111) minutes.
Conveyance self-test routine
recommended polling time: 	 (   2) minutes.
SCT capabilities: 	       (0x3085)	SCT Status supported.

SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   118   099   006    Pre-fail  Always       -       168101376
  3 Spin_Up_Time            0x0003   097   097   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       425
  5 Reallocated_Sector_Ct   0x0033   100   100   010    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   069   060   030    Pre-fail  Always       -       9675211
  9 Power_On_Hours          0x0032   096   096   000    Old_age   Always       -       3982
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       433
183 Runtime_Bad_Block       0x0032   100   100   000    Old_age   Always       -       0
184 End-to-End_Error        0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
188 Command_Timeout         0x0032   100   100   000    Old_age   Always       -       0
189 High_Fly_Writes         0x003a   100   100   000    Old_age   Always       -       0
190 Airflow_Temperature_Cel 0x0022   071   063   045    Old_age   Always       -       29 (Min/Max 20/29)
191 G-Sense_Error_Rate      0x0032   100   100   000    Old_age   Always       -       0
192 Power-Off_Retract_Count 0x0032   100   100   000    Old_age   Always       -       25
193 Load_Cycle_Count        0x0032   100   100   000    Old_age   Always       -       504
194 Temperature_Celsius     0x0022   029   040   000    Old_age   Always       -       29 (0 18 0 0 0)
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       12154757451688
241 Total_LBAs_Written      0x0000   100   253   000    Old_age   Offline      -       14098900823
242 Total_LBAs_Read         0x0000   100   253   000    Old_age   Offline      -       800819281

SMART Error Log Version: 1
No Errors Logged

SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed without error       00%      3005         -
# 2  Extended offline    Completed without error       00%      2008         -
# 3  Extended offline    Completed without error       00%      1014         -
# 4  Extended offline    Completed without error       00%        13         -

SMART Selective self-test log data structure revision number 1
 SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
    1        0        0  Not_testing
    2        0        0  Not_testing
    3        0        0  Not_testing
    4        0        0  Not_testing
    5        0        0  Not_testing
Selective self-test flags (0x0):
  After scanning selected spans, do NOT read-scan remainder of disk.
If Selective self-test is pending on power-up, resume after 0 minute delay.

This is just an example from my current HDD. Technically smartctl -a lists everything, not just attributes, but the whole output is more useful than just the attributes. Some things to note on the output is that SMART support is available and enabled. If it is not available then your device may not support SMART, which can occur if this is an external HDD with a cheap enclosure or if the device is not a HDD/SSD. If it is not enabled, go into your UEFI/BIOS settings and enable it. Also note SMART overall-health self-assessment test result: PASSED, it should be PASSED unless your HDD is failing.

Note the line Auto Offline Data Collection: Enabled, this is a feature that is enabled by default on modern internal HDDs. man smartctl explains what this feature does and how to enable it:

       -o VALUE, --offlineauto=VALUE
              [ATA only] Enables or disables  SMART  automatic  offline  test,
              which  scans  the  drive every four hours for disk defects. This
              command can be given during normal system operation.  The  valid
              arguments to this option are on and off.

This also updates attributes that are marked Offline. Unlike Always updated attributes, Offline attributes are only updated if this is enabled or if you run a SMART test.

Note also that the approximate times for running various tests are listed. We will discuss SMART tests in the next section.

Now about the attributes, their meaning is summarized in man smartctl:

              Each  Attribute  also has a Threshold value (whose range is 0 to
              255) which is printed under the heading "THRESH".  If  the  Nor-
              malized value is less than or equal to the Threshold value, then
              the Attribute is said to have failed.  If  the  Attribute  is  a
              pre-failure Attribute, then disk failure is imminent.

              The Attribute table printed  out  by  smartctl  also  shows  the
              "TYPE"  of  the  Attribute.  Attributes  are one of two possible
              types: Pre-failure or Old age.  Pre-failure Attributes are  ones
              which, if less than or equal to their threshold values, indicate
              pending disk failure.  Old age, or usage  Attributes,  are  ones
              which  indicate end-of-product life from old-age or normal aging
              and wearout, if the Attribute value is less than or equal to the
              threshold.   Please  note: the fact that an Attribute is of type
              'Pre-fail' does not mean that your disk is about  to  fail!   It
              only  has  this  meaning  if  the Attribute´s current Normalized
              value is less than or equal to the threshold value.
              
              If  the  Attribute´s  current  Normalized  value is less than or
              equal to the threshold value, then the "WHEN_FAILED" column will
              display  "FAILING_NOW".  If not, but the worst recorded value is
              less than or equal to the threshold value, then this column will
              display "In_the_past".  If the "WHEN_FAILED" column has no entry
              (indicated by a dash: ´-´) then this Attribute is  OK  now  (not
              failing) and has also never failed in the past.

Thus, the most important attributes are marked Pre-fail. If the value of a Pre-fail attribute is below threshold, the attribute is failing implying that the HDD is failing. A failing attribute will be marked as FAILING_NOW or In_the_past if it has failed now or in the past, respectively. Old_age attribute failures do NOT necessarily mean imminent failure, but rather that the drive is getting old and it should be monitored more carefully or replaced at some point.

For the exact meaning of each attribute, please see the Wiki page. Some specific attributes that I would like to discuss are as follows:

This attribute is important for laptop HDDs, because they default to powering off when not in use. Now, although laptop HDDs are designed to spin up and down more times than desktop HDD and this is an Old_age attribute, it still wears down the drive. Unless you run on batteries all the time you may want to consider turning off this feature by adding this to a boot script such as /etc/rc.d/rc.local:

hdparm -B 254 /dev/sda

There are 3 main types of SMART tests that you can perform.

• short: a superficial test that tests electrical and mechanical performance and updates offline attributes

• conveyance: identifies damage during transport (mostly useful for external or laptop HDDs)

• long: a short test plus it scans the disk surface for bad blocks

These tests are run with the -t option like:

smartctl -t long /dev/sda

These tests can all be run on a running system without major side-effects. If you expect the long test to finish, you should minimize HDD usage as it has to scan the whole disk to finish the test.

After waiting for the test to finish, you can get the results using the -a option as shown in the previous section.

Short and Conveyance tests should always pass. If these fail, check the attributes as the drive is probably failing. A long test can fail if there are bad blocks, and this does NOT mean the drive is failing. The long test stops when it finds an error on the disk, so if there is a bad block it just stops. This doesn’t mean the drive is failing, but you will have to wait for the HDD to remap/reallocate the block, or technically you could try to force it to do so:
http://www.smartmontools.org/browser/trunk/www/badblockhowto.xml
However, this method is difficult to implement safely, so you should usually just wait for the HDD to remap/reallocate.

How often should you run these tests ? That depends. If you run a server then more often is better, the smartmontools site recommends weekly tests. For a home user, I usually run a long test every 1000 power on hours, but that is up to you and also depends on the details of the drive and situation.

A failing drive is defined as:

1. Having a Pre-fail attribute below or near threshold, marked FAILING_NOW or In_the_past.

2. Having an Old_age attribute below or near threshold, marked FAILING_NOW or In_the_past PLUS other signs of failure such as consistent failure of SMART tests, strange noises, slowing down, corrupt data, etc.

A failed long test does NOT mean your drive is failing, it could be just bad blocks. See previous section.

Do not ignore your senses, if the HDD sounds unusual or makes strange noises, monitor it closely and/or replace it. Again, SMART cannot tell you with great accuracy if or when the drive will fail. The drive can fail with above threshold attributes and minimal signs. The only hope you have to keep your data safe is to backup your data, use the 3-2-1 strategy as mentioned above.

• man smartctl