I scanned my hard drive with HD Tune and I'm getting the following data:

HD Tune: ST9120817AS Health

ID                               Current  Worst    ThresholdData        Status   
(01) Raw Read Error Rate         100      253      6        0           Ok       
(03) Spin Up Time                100      99       85       0           Ok       
(04) Start/Stop Count            1        1        20       579275      Failed   
(05) Reallocated Sector Count    100      100      36       0           Ok       
(07) Seek Error Rate             83       60       30       201641351   Ok       
(09) Power On Hours Count        90       90       0        9064        Ok       
(0A) Spin Retry Count            100      100      34       0           Ok       
(0C) Power Cycle Count           96       96       20       4456        Ok       
(BB) (unknown attribute)         100      100      0        0           Ok       
(BD) (unknown attribute)         98       98       0        2           Ok       
(BE) Airflow Temperature         56       39       45       1023606828  Ok       
(BF) G-sense Error Rate          100      100      0        43          Ok       
(C0) Power Off Retract Count     100      100      0        41          Ok       
(C1) Load Cycle Count            1        1        0        576186      Ok       
(C2) Temperature                 44       61       0        44          Ok       
(C3) Hardware ECC Recovered      64       59       0        148606121   Ok       
(C4) Reallocated Event Count     93       93       0        6385        Ok       
(C5) Current Pending Sector      100      100      0        0           Ok       
(C6) Offline Uncorrectable       100      100      0        0           Ok       
(C7) Ultra DMA CRC Error Count   200      200      0        0           Ok       
(C8) Write Error Rate            100      253      0        0           Ok       
(CA) TA Counter Increased        100      253      0        0           Ok       
(F0) Head Flying Hours           0        0        0        0           Ok       
(F1) (unknown attribute)         0        0        0        1419893233  Ok       
(F2) (unknown attribute)         0        0        0        239964343   Ok       

Power On Time         : 9064
Health Status         : Failed

I am particularly worried about the start/stop count. What does the Data value mean? Is that the actual number of times the disk started? That would be very high since my other drive which is just as old has only 1433. It's so high that I'm wondering if I can even trust these values, especially since everything seems to work fine. How likely is it that I'm getting false readings?

How afraid should I be of losing all my data?

  • Is this on a so-called green drive? Those are often configured to spin down on their own after a very short period to safe power. This makes sense in some configurations (e.g. on an external drive containing movies), but plays havoc in when used on drives containing log files. – Hennes Jul 3 '12 at 15:52

You should always have all of it backed up. That said, SMART is not so reliable as an indicator of the drive malfunctioning, so don't be too worried.

  • @suandos The disk is only used for VM's, data is in online repo's. reïnstalling software is a pain though. PS: I'm still too worried :) – Renaat De Muynck Jul 4 '12 at 10:47

Yes, the start/stop value should show the number of times the spindle motor has been started/stopped and is usually a reliable value. This would imply that the motor has stopped/started about each minute. Perhaps you used an inappropriate low timeout for the drive to enter standby or a really aggressive APM setting. If you could boot a Linux system the output of hdparm -B /dev/sda (assuming your HDD is sda) would be interesting. You might also want to use a better tool like smartmontools to check your S.M.A.R.T. values, since your tool fails to identify/decode some attributes for unknown reasons. smartctl -a /dev/sda would be the command to run. But since there are no known firmware bugs for your drive that could give wrong results without workarounds, the already known data will probably be the same. smartctl -a however has more output, so there might be some additional interesting data.

The value for Reallocated Event Count is also interesting. It says your drive tried to reallocate damaged sectors 6385 times but it looks like it failed each time since Reallocated Sector Count is still 0. I would worry about that. I recommend to backup as soon as possible, to execute a long test on the drive using smartctl -t long /dev/sda and then repeat smartctl -a /dev/sda after the given polling time. If the test result is passed, you might be able to continue using the drive but you should be prepared for a higher risk of failure.

If you prefer Windows/a GUI you could use GSmartControl instead of smartmontools to view data/execute the test.

  • Thanks for the clear explanation, it gave me some insight. I'm running Windows 7, so I'll try GSmartControl. I already did an extended error scan with HD Tune with no errors. I don't know if this is related but about once a week or so my system freezes and my hard drive led stays on. I have to manually shut down my laptop (I think it could be a loose cable but I'm really guessing here). Could this have caused the high start/stop count? Could a failing/disconnected HD freeze up my OS if it's on another drive? – Renaat De Muynck Jul 4 '12 at 10:24
  • Your freezes are likely related to your drive if the LED is always on when it freezes. Have you tried to wait for some time if this happens? It might continue after some minutes. You should also check your S.M.A.R.T. error log with GSmartControl/smartctl. A loose cable is unlikely as there is none in normal laptop computers. You still might want to re-seat the disk to make sure it is connected firmly. If you had no serious problem reading from/writing to the disk, the high start/stop count is probably not caused by bad connection. I rather suspect a extremely aggressive APM setting. – Gurken Papst Jul 4 '12 at 18:21
  • Check that (as already recommended) with hdparm. You can use a Linux live system for that. Reading from/writing to a damaged disk can lock your whole system for minutes, until a timeout occurs. So yes, this can happen if your OS is not on that drive. A completely disconnected drive however will of course not cause any freezes, but I cannot see how this could be related to your problem. – Gurken Papst Jul 4 '12 at 18:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.