S.M.A.R.T. – Wikipedia

S.M.A.R.T. (akronym för Self-Monitoring Analysis and Reporting Technology) är en funktion i vissa hårddiskar som gör det möjligt uppskatta tillståndet på dessa. Diagnostiken som görs varierar mellan olika hårddiskar, men innefattar ofta saker som att söktid och andel felläsningar ligger inom ett acceptabelt intervall. Antalet timmar hårddisken varit påslagen, antalet gånger den startat upp, och den högsta temperatur den blivit utsatt för är andra vanliga parametrar som kan läsas ut.

Vissa datorer visar en varning vid uppstart om dessa parametrar indikerar att något verkar fel. I sådana fall bör en backup omedelbart göras och hårddisken bytas ut då den har en klart ökad risk för att inom kort helt gå sönder och bli oläsbar.

En fältstudie på Google^[1] som täckte över 100 000 enheter för konsumentklass från december 2005 till augusti 2006 hittade följande samband mellan viss SMART-information och årliga hårddiskhaverier:

• Under 60 dagar efter det första okorrigerbara felet på en enhet (SMART-data 198/0xC6) som upptäcktes genom en offline-skanning, var enheten 39 gånger i genomsnitt mer benägen att haverera, än en felfri enhet.
• Första registreringen av SMART-data för omlokalisering (5/0x05 och 196/0xC5) samt väntad omlokalisering (SMART-data 197/0xC5) var också starkt korrelerad med högre sannolikhet för haveri.
• Små samband för ökad temperatur och inget samband för användningsnivå.
• 56% av de havererade enheterna hade inget registrerat i något av de "fyra starka S.M.A.R.T.-varningarna" (läsfel eller olika typer av omlokaliseringsräknare).
• 36% av de havererade enheterna hade inga S.M.A.R.T.-fel alls, utom temperaturen, vilket innebär att S.M.A.R.T.-data ensamt bara ger begränsad möjlighet att förutse haveri.^[2]

Varje tillverkare definierar en uppsättning attribut^[3][4], och sätter olika tröskelvärden för attribut som inte ska passera under normal drift. Varje attribut har ett råvärde(RAW) i decimal- eller hexadecimal form vars betydelse bestäms av drivenhetstillverkaren. Dessa råvärden står för antal gånger något har är inträffat, drifttimmar, uppstarter, temperatur, söktider, time-outer, läsfel, överförda datamängder, etc. Viktigare är att avläsa aktuellt attributvärde i förhållande till sämsta uppmätta värde och tröskelvärde, ofta uttryckt i skalor 252→0, 200→0 eller 100→0, där högt aktuellt värde är bäst.^[5]

Tillverkare som har implementerat minst ett S.M.A.R.T.-attribut i olika produkter inkluderar Samsung, Seagate, IBM (Hitachi), Fujitsu, Maxtor, Toshiba, Intel, sTec, Inc., Western Digital och ExcelStor Technology.

Följande diagram visar några S.M.A.R.T.-attribut och den typiska betydelsen av deras råvärden. Normaliserade värden kartläggs vanligtvis så att de högre värdena är bättre (undantag inkluderar körtemperatur, antal läshuvudbelastningar-/avlastningscykler^[6]), men högre RAW-värden kan vara bättre eller sämre beroende på attributet och tillverkaren. Till exempel attributet "Reallocated Sectors Count/Antal omlokaliserade sektorer" aktuella normaliserade värde minskar när antalet omlokaliserade sektorer ökar. I detta fall indikerar attributets "RAW"-värde ofta antalet sektorer som har omfördelats, även om leverantörer inte på något sätt säger sig följa denna konvention.

Eftersom tillverkare inte nödvändigtvis håller med om exakta attributdefinitioner och mätenheter, är följande lista med attribut bara en allmän guide.

Enheter stöder inte alla attributkoder (ibland förkortade "ID", för "identifierare", i tabeller). Vissa koder är specifika för specifika drivenheter (magnetisk tallrik, blixt, SSD). Enheter kan använda olika koder för samma parameter, se till exempel koder 193 och 225.

Legend/Teckenförklaring

ID	193 0xC1	Attribut i decimal och hexadecimal form
Önskvärt	Hög	Högre RAW-värde är bättre.
	Låg	Lägre RAW-värde är bättre.
! (Kritisk)		Kritiskt attribut. Specifika värden kan förutse enhetshaveri.

ID	Attribut	Önskvärt	!	Beskrivning
01 0x01	Read Error Rate	Låg		(Vendor specific raw value.) Stores data related to the rate of hardware read errors that occurred when reading data from a disk surface. The raw value has different structure for different vendors and is often not meaningful as a decimal number.
02 0x02	Throughput Performance	Hög		Overall (general) throughput performance of a hard disk drive. If the value of this attribute is decreasing there is a high probability that there is a problem with the disk.
03 0x03	Spin-Up Time	Låg		Average time of spindle spin up (from zero RPM to fully operational [milliseconds]).
04 0x04	Start/Stop Count			A tally of spindle start/stop cycles. The spindle turns on, and hence the count is increased, both when the hard disk is turned on after having before been turned entirely off (disconnected from power source) and when the hard disk returns from having previously been put to sleep mode.[7]
05 0x05	Reallocated Sectors Count	Låg	[8][9][10]	Count of reallocated sectors. The raw value represents a count of the bad sectors that have been found and remapped.[11] Thus, the higher the attribute value, the more sectors the drive has had to reallocate. This value is primarily used as a metric of the life expectancy of the drive; a drive which has had any reallocations at all is significantly more likely to fail in the immediate months.[8][12]
06 0x06	Read Channel Margin			Margin of a channel while reading data. The function of this attribute is not specified.
07 0x07	Seek Error Rate	Mall:Varies		(Vendor specific raw value.) Rate of seek errors of the magnetic heads. If there is a partial failure in the mechanical positioning system, then seek errors will arise. Such a failure may be due to numerous factors, such as damage to a servo, or thermal widening of the hard disk. The raw value has different structure for different vendors and is often not meaningful as a decimal number.
08 0x08	Seek Time Performance	Hög		Average performance of seek operations of the magnetic heads. If this attribute is decreasing, it is a sign of problems in the mechanical subsystem.
09 0x09	Power-On Hours			Count of hours in power-on state. The raw value of this attribute shows total count of hours (or minutes, or seconds, depending on manufacturer) in power-on state.[13] "By default, the total expected lifetime of a hard disk in perfect condition is defined as 5 years (running every day and night on all days). This is equal to 1825 days in 24/7 mode or 43800 hours."[14] On some pre-2005 drives, this raw value may advance erratically and/or "wrap around" (reset to zero periodically).[15]
10 0x0A	Spin Retry Count	Låg	[16]	Count of retry of spin start attempts. This attribute stores a total count of the spin start attempts to reach the fully operational speed (under the condition that the first attempt was unsuccessful). An increase of this attribute value is a sign of problems in the hard disk mechanical subsystem.
11 0x0B	Recalibration Retries or Calibration Retry Count	Låg		This attribute indicates the count that recalibration was requested (under the condition that the first attempt was unsuccessful). An increase of this attribute value is a sign of problems in the hard disk mechanical subsystem.
12 0x0C	Power Cycle Count			This attribute indicates the count of full hard disk power on/off cycles.
13 0x0D	Soft Read Error Rate	Låg		Uncorrected read errors reported to the operating system.
22 0x16	Current Helium Level			Specific to He8 drives from HGST. This value measures the helium inside of the drive specific to this manufacturer. It is a pre-fail attribute that trips once the drive detects that the internal environment is out of specification.[17]
170 0xAA	Available Reserved Space			See attribute E8.[18]
171 0xAB	SSD Program Fail Count			(Kingston) The total number of flash program operation failures since the drive was deployed.[19] Identical to attribute 181.
172 0xAC	SSD Erase Fail Count			(Kingston) Counts the number of flash erase failures. This attribute returns the total number of Flash erase operation failures since the drive was deployed. This attribute is identical to attribute 182.
173 0xAD	SSD Wear Leveling Count			Counts the maximum worst erase count on any block.
174 0xAE	Unexpected power loss count			Also known as "Power-off Retract Count" per conventional HDD terminology. Raw value reports the number of unclean shutdowns, cumulative over the life of an SSD, where an "unclean shutdown" is the removal of power without STANDBY IMMEDIATE as the last command (regardless of PLI activity using capacitor power). Normalized value is always 100.[20]
175 0xAF	Power Loss Protection Failure			Last test result as microseconds to discharge cap, saturated at its maximum value. Also logs minutes since last test and lifetime number of tests. Raw value contains the following data: Bytes 0-1: Last test result as microseconds to discharge cap, saturates at max value. Test result expected in range 25 <= result <= 5000000, lower indicates specific error code. Bytes 2-3: Minutes since last test, saturates at max value. Bytes 4-5: Lifetime number of tests, not incremented on power cycle, saturates at max value. Normalized value is set to one on test failure or 11 if the capacitor has been tested in an excessive temperature condition, otherwise 100.[20]
176 0xB0	Erase Fail Count			S.M.A.R.T. parameter indicates a number of flash erase command failures.[21]
177 0xB1	Wear Range Delta			Delta between most-worn and least-worn Flash blocks. It describes how good/bad the wearleveling of the SSD works on a more technical way.
179 0xB3	Used Reserved Block Count Total			"Pre-Fail" attribute used at least in Samsung devices.
180 0xB4	Unused Reserved Block Count Total			"Pre-Fail" attribute used at least in HP devices.
181 0xB5	Program Fail Count Total or Non-4K Aligned Access Count	Låg		Total number of Flash program operation failures since the drive was deployed.[22] Number of user data accesses (both reads and writes) where LBAs are not 4 KiB aligned (LBA % 8 != 0) or where size is not modulus 4 KiB (block count != 8), assuming logical block size (LBS) = 512 B.[23]
182 0xB6	Erase Fail Count			"Pre-Fail" Attribute used at least in Samsung devices.
183 0xB7	SATA Downshift Error Count or Runtime Bad Block	Låg		Western Digital, Samsung or Seagate attribute: Either the number of downshifts of link speed (e.g. from 6Gbps to 3Gbps) or the total number of data blocks with detected, uncorrectable errors encountered during normal operation.[24] Although degradation of this parameter can be an indicator of drive aging and/or potential electromechanical problems, it does not directly indicate imminent drive failure.[25]
184 0xB8	End-to-End error / IOEDC	Låg	[26]	This attribute is a part of Hewlett-Packard's SMART IV technology, as well as part of other vendors' IO Error Detection and Correction schemas, and it contains a count of parity errors which occur in the data path to the media via the drive's cache RAM.[27]
185 0xB9	Head Stability			Western Digital attribute.
186 0xBA	Induced Op-Vibration Detection			Western Digital attribute.
187 0xBB	Reported Uncorrectable Errors	Låg	[28]	The count of errors that could not be recovered using hardware ECC (see attribute 195).[28]
188 0xBC	Command Timeout	Låg	[29]	The count of aborted operations due to HDD timeout. Normally this attribute value should be equal to zero.[29]
189 0xBD	High Fly Writes	Låg		HDD manufacturers implement a flying height sensor that attempts to provide additional protections for write operations by detecting when a recording head is flying outside its normal operating range. If an unsafe fly height condition is encountered, the write process is stopped, and the information is rewritten or reallocated to a safe region of the hard drive. This attribute indicates the count of these errors detected over the lifetime of the drive. This feature is implemented in most modern Seagate drives[30] and some of Western Digital's drives, beginning with the WD Enterprise WDE18300 and WDE9180 Ultra2 SCSI hard drives, and will be included on all future WD Enterprise products.[31]
190 0xBE	Temperature Difference or Airflow Temperature	Mall:Varies		Value is equal to (100-temp. °C), allowing manufacturer to set a minimum threshold which corresponds to a maximum temperature. This also follows the convention of 100 being a best-case value and lower values being undesirable. However, some older drives may instead report raw Temperature (identical to 0xC2) or Temperature minus 50 here.
191 0xBF	G-sense Error Rate	Låg		The count of errors resulting from externally induced shock and vibration.
192 0xC0	Power-off Retract Count, Emergency Retract Cycle Count (Fujitsu),[32] or Unsafe Shutdown Count	Låg		Number of power-off or emergency retract cycles.[5][33]
193 0xC1	Load Cycle Count or Load/Unload Cycle Count (Fujitsu)	Låg		Count of load/unload cycles into head landing zone position.[32] Some drives use 225 (0xE1) for Load Cycle Count instead. Western Digital rates their VelociRaptor drives for 600,000 load/unload cycles,[34] and WD Green drives for 300,000 cycles;[35] the latter ones are designed to unload heads often to conserve power. On the other hand, the WD3000GLFS (a desktop drive) is specified for only 50,000 load/unload cycles.[36] Some laptop drives and "green power" desktop drives are programmed to unload the heads whenever there has not been any activity for a short period, to save power.[37][38] Operating systems often access the file system a few times a minute in the background,[39] causing 100 or more load cycles per hour if the heads unload: the load cycle rating may be exceeded in less than a year.[40] There are programs for most operating systems that disable the Advanced Power Management (APM) and Automatic acoustic management (AAM) features causing frequent load cycles.[41][42]
194 0xC2	Temperature or Temperature Celsius	Låg		Indicates the device temperature, if the appropriate sensor is fitted. Lowest byte of the raw value contains the exact temperature value (Celsius degrees).[43]
195 0xC3	Hardware ECC Recovered	Mall:Varies		(Vendor-specific raw value.) The raw value has different structure for different vendors and is often not meaningful as a decimal number.
196 0xC4	Reallocation Event Count[32]	Låg	[1]	Count of remap operations. The raw value of this attribute shows the total count of attempts to transfer data from reallocated sectors to a spare area. Both successful and unsuccessful attempts are counted.[44]
197 0xC5	Current Pending Sector Count[32]	Låg	[1][26][29]	Count of "unstable" sectors (waiting to be remapped, because of unrecoverable read errors). If an unstable sector is subsequently read successfully, the sector is remapped and this value is decreased. Read errors on a sector will not remap the sector immediately (since the correct value cannot be read and so the value to remap is not known, and also it might become readable later); instead, the drive firmware remembers that the sector needs to be remapped, and will remap it the next time it's written.[45] However, some drives will not immediately remap such sectors when written; instead the drive will first attempt to write to the problem sector and if the write operation is successful then the sector will be marked good (in this case, the "Reallocation Event Count" (0xC4) will not be increased). This is a serious shortcoming, for if such a drive contains marginal sectors that consistently fail only after some time has passed following a successful write operation, then the drive will never remap these problem sectors.
198 0xC6	(Offline) Uncorrectable Sector Count[32]	Låg	[1][8]	The total count of uncorrectable errors when reading/writing a sector. A rise in the value of this attribute indicates defects of the disk surface and/or problems in the mechanical subsystem.[1][29][26]
199 0xC7	UltraDMA CRC Error Count	Låg		The count of errors in data transfer via the interface cable as determined by ICRC (Interface Cyclic Redundancy Check).
200 0xC8	Multi-Zone Error Rate [46]	Låg		The count of errors found when writing a sector. The higher the value, the worse the disk's mechanical condition is.
200 0xC8	Write Error Rate (Fujitsu)	Låg		The total count of errors when writing a sector.[47]
201 0xC9	Soft Read Error Rate or TA Counter Detected	Låg	[48]	Count indicates the number of uncorrectable software read errors.[48]
202 0xCA	Data Address Mark errors or TA Counter Increased	Låg		Count of Data Address Mark errors (or vendor-specific).[5]
203 0xCB	Run Out Cancel	Låg		The number of errors caused by incorrect checksum during the error correction.
204 0xCC	Soft ECC Correction	Låg		Count of errors corrected by the internal error correction software.[5]
205 0xCD	Thermal Asperity Rate	Low		Count of errors due to high temperature.[49]
206 0xCE	Flying Height			Height of heads above the disk surface. If too low, head crash is more likely; if too high, read/write errors are more likely.[5][50]
207 0xCF	Spin High Current	Låg		Amount of surge current used to spin up the drive.[49]
208 0xD0	Spin Buzz			Count of buzz routines needed to spin up the drive due to insufficient power.[49]
209 0xD1	Offline Seek Performance			Drive’s seek performance during its internal tests.[49]
210 0xD2	Vibration During Write			Found in Maxtor 6B200M0 200GB and Maxtor 2R015H1 15GB disks.
211 0xD3	Vibration During Write			A recording of a vibration encountered during write operations.[51]
212 0xD4	Shock During Write			A recording of shock encountered during write operations.[19][52]
220 0xDC	Disk Shift	Låg		Distance the disk has shifted relative to the spindle (usually due to shock or temperature). Unit of measure is unknown.[19]
221 0xDD	G-Sense Error Rate	Låg		The count of errors resulting from externally induced shock and vibration.
222 0xDE	Loaded Hours			Time spent operating under data load (movement of magnetic head armature).[19]
223 0xDF	Load/Unload Retry Count			Count of times head changes position.[19]
224 0xE0	Load Friction	Låg		Resistance caused by friction in mechanical parts while operating.[19]
225 0xE1	Load/Unload Cycle Count	Låg		Total count of load cycles[19] Some drives use 193 (0xC1) for Load Cycle Count instead. See Description for 193 for significance of this number.
226 0xE2	Load 'In'-time			Total time of loading on the magnetic heads actuator (time not spent in parking area).[19]
227 0xE3	Torque Amplification Count	Låg		Count of attempts to compensate for platter speed variations.[53]
228 0xE4	Power-Off Retract Cycle	Låg		The number of power-off cycles which are counted whenever there is a "retract event" and the heads are loaded off of the media such as when the machine is powered down, put to sleep, or is idle.[5][33]
230 0xE6	GMR Head Amplitude (magnetic HDDs), Drive Life Protection Status (SSDs)			Amplitude of "thrashing" (repetitive head moving motions between operations).[5][54] In solid-state drives, indicates whether usage trajectory is outpacing the expected life curve[55]
231 0xE7	Life Left (SSDs) or Temperature			Indicates the approximate SSD life left, in terms of program/erase cycles or available reserved blocks.[55] A normalized value of 100 represents a new drive, with a threshold value at 10 indicating a need for replacement. A value of 0 may mean that the drive is operating in read-only mode to allow data recovery.[56] Previously (pre-2010) occasionally used for Drive Temperature (more typically reported at 0xC2).
232 0xE8	Endurance Remaining or Available Reserved Space			Number of physical erase cycles completed on the SSD as a percentage of the maximum physical erase cycles the drive is designed to endure. Intel SSDs report the available reserved space as a percentage of the initial reserved space.
233 0xE9	Media Wearout Indicator (SSDs) or Power-On Hours			Intel SSDs report a normalized value from 100, a new drive, to a minimum of 1. It decreases while the NAND erase cycles increase from 0 to the maximum-rated cycles. Previously (pre-2010) occasionally used for Power-On Hours (more typically reported in 0x09).
234 0xEA	Average erase count AND Maximum Erase Count			Decoded as: byte 0-1-2 = average erase count (big endian) and byte 3-4-5 = max erase count (big endian).[57]
235 0xEB	Good Block Count AND System(Free) Block Count			Decoded as: byte 0-1-2 = good block count (big endian) and byte 3-4 = system (free) block count.
240 0xF0	Head Flying Hours or 'Transfer Error Rate' (Fujitsu)			Time spent during the positioning of the drive heads.[5][58] Some Fujitsu drives report the count of link resets during a data transfer.[59]
241 0xF1	Total LBAs Written			Total count of LBAs written.
242 0xF2	Total LBAs Read			Total count of LBAs read. Some S.M.A.R.T. utilities will report a negative number for the raw value since in reality it has 48 bits rather than 32.
243 0xF3	Total LBAs Written Expanded			The upper 5 bytes of the 12-byte total number of LBAs written to the device. The lower 7 byte value is located at attribute 0xF1.[60]
244 0xF4	Total LBAs Read Expanded			The upper 5 bytes of the 12-byte total number of LBAs read from the device. The lower 7 byte value is located at attribute 0xF2.[60]
249 0xF9	NAND Writes (1GiB)			Total NAND Writes. Raw value reports the number of writes to NAND in 1 GB increments.[61]
250 0xFA	Read Error Retry Rate	Låg		Count of errors while reading from a disk.[19]
251 0xFB	Minimum Spares Remaining			The Minimum Spares Remaining attribute indicates the number of remaining spare blocks as a percentage of the total number of spare blocks available.[62]
252 0xFC	Newly Added Bad Flash Block			The Newly Added Bad Flash Block attribute indicates the total number of bad flash blocks the drive detected since it was first initialized in manufacturing.[62]
254 0xFE	Free Fall Protection	Låg		Count of "Free Fall Events" detected.[63]

Tröskelvärdedatum (TEC) är ett bedömt datum då ett kritiskt enhetsstatistikattribut når sitt tröskelvärde. När "Drive Health"-programvaran rapporterar en "närmaste TEC" bör den betraktas som ett "Failure date", dvs. då enheter havererar. Ibland anges inget datum och enheten kan förväntas fungera utan fel.^[64]

För att förutsäga datum följer enheten hastigheten som attributet ändras i. Observera att TEC-datum endast är uppskattningar; hårddiskar kan haverera både mycket tidigare eller mycket senare.^[65]

SMART-enheter kan erbjuda flera självtester:^[66][67][68]

Kortfattad (Short)

Kontrollerar den elektriska och mekaniska prestandan såväl som skivans läsprestanda. Elektriska tester kan inkludera ett test av buffert-RAM, ett test av läs-/skrivkrets eller ett test av läs-/skrivhuvud. Mekaniska tester inkluderar sökning och motorstyrning, genom att skanna delar av enhetens yta. Vidare kontrolleras listan över sektorer som kan ha läsfel. Testet utformas olika av olika tillverkare och det körs under en begränsad tid, vanligtvis under två minuter.

Utförlig (Long/Extended)

En längre och mer grundlig version av det korta självtestet som skannar hela skivytan utan tidsgräns. Detta test tar vanligtvis flera timmar, beroende på läs-/skrivhastighet på enheten och dess storlek.

Kortfattad tillverkaranpassad (Conveyance)

Avsedd som ett snabbtest för att identifiera skador som uppstod vid transport av enheten från enhetens tillverkare till datortillverkaren.^[69]

Endast tillgängligt på ATA-enheter och tar vanligtvis flera minuter.

Selektivt (Selective)

Vissa enheter tillåter selektivt självtest på bara en del av ytan.^[70]

Självtestloggarna för SCSI- och ATA-enheter är något olika. Det är möjligt för det långa testet att passera även om det korta testet misslyckas.^[71]

Diskens självtestlogg kan innehålla upp till 21 skrivskyddade poster. När loggen är fylld tas gamla poster bort.^[72]

• Seagate/Samsung
• Western Digital/HGST

Den här artikeln är helt eller delvis baserad på material från engelskspråkiga Wikipedia, S.M.A.R.T., 31 augusti 2019.

• Computer World - Fem SMART-data som förutspår hårddiskhaveri