Reliability of S.M.A.R.T. Selftest

Question

I let smartd check my hard drives regularly. Recently I have been notified about a failed short selftest on one of my hard-drivers.

 smartctl 6.4 2014-10-07 r4002 [x86_64-linux-3.16.0-0.bpo.4-amd64] (local build)
    Copyright (C) 2002-14, Bruce Allen, Christian Franke, www.smartmontools.org

    === START OF INFORMATION SECTION ===
    Model Family:     Western Digital RE4
    Device Model:     WDC WD1003FBYX-01Y7B0
    Serial Number:    WD-WCAW31677053
    LU WWN Device Id: 5 0014ee 25b0c013e
    Firmware Version: 01.01V01
    User Capacity:    1.000.204.886.016 bytes [1,00 TB]
    Sector Size:      512 bytes logical/physical
    Rotation Rate:    7200 rpm
    Device is:        In smartctl database [for details use: -P show]
    ATA Version is:   ATA8-ACS (minor revision not indicated)
SATA Version is:  SATA 2.6, 3.0 Gb/s
Local Time is:    Mon Dec  7 13:13:22 2015 CET
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:  (0x85) Offline data collection activity
                                        was aborted by an interrupting command from host.
                                        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:                (16680) 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:        (   2) minutes.
Extended self-test routine
recommended polling time:        ( 172) minutes.
Conveyance self-test routine
recommended polling time:        (   5) minutes.
SCT capabilities:              (0x303f) SCT Status supported.
                                        SCT Error Recovery Control supported.
                                        SCT Feature Control supported.
                                        SCT Data Table supported.

SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x002f   200   200   051    Pre-fail  Always       -       1
  3 Spin_Up_Time            0x0027   176   173   021    Pre-fail  Always       -       4191
  4 Start_Stop_Count        0x0032   100   100   000    Old_age   Always       -       37
  5 Reallocated_Sector_Ct   0x0033   200   200   140    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x002e   200   200   000    Old_age   Always       -       0

      9 Power_On_Hours          0x0032   070   070   000    Old_age   Always       -       21991
     10 Spin_Retry_Count        0x0032   100   253   000    Old_age   Always       -       0
     11 Calibration_Retry_Count 0x0032   100   253   000    Old_age   Always       -       0
     12 Power_Cycle_Count       0x0032   100   100   000    Old_age   Always       -       37
    192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       35
    193 Load_Cycle_Count        0x0032   200   200   000    Old_age   Always       -       1
    194 Temperature_Celsius     0x0022   110   088   000    Old_age   Always       -       37
    196 Reallocated_Event_Count 0x0032   200   200   000    Old_age   Always       -       0
    197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       0
    198 Offline_Uncorrectable   0x0030   200   200   000    Old_age   Offline      -       0
    199 UDMA_CRC_Error_Count    0x0032   200   200   000    Old_age   Always       -       0
    200 Multi_Zone_Error_Rate   0x0008   200   200   000    Old_age   Offline      -       0

    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  Short offline       Completed without error       00%     21981         -
    # 2  Extended offline    Completed without error       00%     21972         -
    # 3  Short offline       Completed without error       00%     21963         -
    # 4  Extended offline    Completed: read failure       90%     21934         4631341
    # 5  Short offline       Completed: read failure       50%     21933         4631341
    # 6  Short offline       Completed without error       00%     21909         -
    # 7  Short offline       Completed without error       00%     21885         -
    # 8  Short offline       Completed without error       00%     21861         -
    # 9  Short offline       Completed without error       00%     21837         -
    #10  Short offline       Completed without error       00%     21813         -
    #11  Short offline       Completed without error       00%     21789         -
    #12  Extended offline    Completed without error       00%     21769         -
    #13  Short offline       Completed without error       00%     21765         -
    #14  Short offline       Completed without error       00%     21742         -
    #15  Short offline       Completed without error       00%     21718         -
    #16  Short offline       Completed without error       00%     21694         -
    #17  Short offline       Completed without error       00%     21670         -
    #18  Short offline       Completed without error       00%     21646         -
    #19  Short offline       Completed without error       00%     21622         -
    #20  Extended offline    Completed without error       00%     21602         -
    #21  Short offline       Completed without error       00%     21598         -
    2 of 2 failed self-tests are outdated by newer successful extended offline self-test # 2

    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.

I then started an extended offline self-test and it also aborted because of a read failure on the same LBA. I then removed the disk from mdadm-raid and started a destructive test with badblocks - it passed three times and surprisingly no errors were found. After another self-test the log error count decreased to 0.

Is my drive cheating on me? (It's a Western Digital RE4.) It reports no Current_Pending_Sector, Offline_Uncorrectable or Reallocated_Event_Count. It reports 1 at Raw_Read_Error_Rate and it reported 1 for Multi_Zone_Error_Rate but this is back at 0 now.

I will replace the drive, but I am not sure if I should overthink my assumption that an extended offline self-test is reliable.

Is the test reliable **at what**? As the famous google paper made clear, if SMART says your drive is going to fail, it almost certainly will fail, and soon; if SMART says your drive is OK, that doesn't mean it's *not* going to fail. See eg [this answer](http://serverfault.com/a/223043/55514) for more detail. — MadHatter, Dec 07 '15 at 11:28
I know that google paper, the question is how reliable is the output regarding the read errors in the past and the number of reallocated sectors and if a succeeding Extended offline test indicates that any single existing sector on the disk can be read. The problem is it first said it will fail and now seems to have forgotten it. The drive WILL be replaced, I'm just curios. I never experienced a succeeding self-test after a failed one. /deleted old comment because my previous answer could lead to misunderstandings — Christian, Dec 07 '15 at 11:50
`2 of 2 failed self-tests are outdated by newer successful extended offline self-test # 2` — Michael Hampton, Dec 07 '15 at 14:16
Yes, Michael, exactly that's what I don't understand, how the result of an extended self-test could change. I doubt that this is reliable as it failed in the past. I thought it would possibly only pass again in case sectors are reallocated but at the same time no sectors were reallocated. SMART seems to be reporting something false what it should "know" better. — Christian, Dec 07 '15 at 14:53
That's my point about the google paper. Once you realise that a *failed* test indicates *dire doom*, but a successful test tells you *nothing*, it isn't odd that a failed test is followed by successful ones. The successful ones don't contradict the failed ones, because they aren't telling you that everything's all right: they're merely telling that you, in a single three-hour test (short tests are worse than useless) nothing heinous could be found. You have a record that it once was, and you're doing the sane thing, replacing the disc. — MadHatter, Dec 07 '15 at 15:44
I see, but data is not only useless, also false. How can LBA 4631341 be not errourneous anymore while no blocks have been reallocated. S.M.A.R.T. seems to be very D.U.M.B. here. But facing the fact, that three complete wipes with badblocks did not find any problem it seems that negative results are not reliable, too. — Christian, Dec 07 '15 at 15:54
Are you seriously asking if a read failure can be intermittent? Because yes, it can. What you're seeing indicates flaky hardware, to me, and you're doing what we should all do with flaky hardware: skipping it. SMART is only seeming dumb to you because you are expecting it, in three hours, and without testing writes, to give you a meaningful test of the current capabilities and the future health of a disc. — MadHatter, Dec 07 '15 at 16:00
Sorry for being dumb, I just never have seen a drive where an extended selftest reports the same LBA as corrupt several times and succeed later on the same, and I've examined a lot of hard drives when I worked as a technician in a hosting providers datacenter for several years. It's just something seems odd here. I have never seen a drive where SMART reports errors while a triple full wipe with badblocks don't while at the same time smart reports 0 reallocated sectors. — Christian, Dec 07 '15 at 16:17
I just wonder if the disk has done some magic to reallocate a block without SMART knowing...if there are probably some vendor-specific technology I might not be aware of... what effectively makes SMART more unreliable. Again, sorry for being dumb. — Christian, Dec 07 '15 at 16:18

score 3 · Answer 1 · answered Dec 07 '15 at 11:29

3

Does your SMART status now report reallocated sectors? You did a destructive test, which writes. When the drive finds sectors it can't write, it will reallocate them. Keeping an eye on the change of this SMART value is important.

Generally, SMART more often does not see errors than it makes up errors. Self tests and reallocated sectors are the more reliable parameters, so your drive is probably 'busting' (on it's way to being 'busted' :)).

answered Dec 07 '15 at 11:29

Halfgaar

7,921
5
42
81

This is what I expected to happen when running badblocks over the whole drive - but it didn't. No reallocated sectors and no pending sectors after full badblocks run over two days. I'll add more smartctl output. And selftest also does not find the broken LBA anymore. That's why I am confused and doubt what my hard-drive is telling me :D – Christian Dec 07 '15 at 11:39

Reliability of S.M.A.R.T. Selftest

1 Answers1