So I bought a "Asus X540LA Laptop, Intel Core i7-5500U 2.4GHz, 8GB RAM, 1TB HDD" in June this year. And today I got a warning that the HDD would fail in <24 hours. I ran smartctl to diagnose the issue and got the following output (on pastebin because it's quite large) https://pastebin.com/xrQp4hpv

Is this a legitimate issue? Is there anything I can do?

  • What gave you the warning, and any indication (or in it's docs) why? – Xen2050 Dec 22 '17 at 1:57
  • GNOME notification popped up saying it would fail soon – user1646196 Dec 22 '17 at 11:12
  • Probably some smartctl daemon monitor always checking I suppose – Xen2050 Dec 22 '17 at 20:25

I would first make a backup (it's always a good decision) just to sleep better at night.

That being said, from what I can tell by reading the output those 3 errors are quite old:

Error 2 [1] occurred at disk power-on lifetime: 1295 hours (53 days + 23 hours)

It's possible that you had a power failure 53 days + 23 hours ago (check the messages log) which caused those errors to be logged and the disk has no actual issue.


You got the warning due to a high Raw_Read_Error_Rate

 1 Raw_Read_Error_Rate     PO-R--   062   062   062    NOW  572391424

You also have 3 logged crc errors with the sector number given (ABRT at LBA = 0x6e44b467 = 1849996391), but perhaps they all occurred at the same time due to mechanical or electrical disturbance, and may not be associated with the failure.

The read error rate however suggests that many blocks are having to be read several times by this drive until it gets a good crc that means the data was read ok.

What you can do is run a short or long non-destructive selftest that will try to read each cylinder or sector and provide statistics on how successful it was. Do smartctl -t short /dev/sda or smartctl -t long /dev/sda and it will say something like Please wait 180 seconds for test to complete. After this time, redo the smartctl -a to see the result and any new information it provides.

  • So that's what the self-tests do, just read data? I'm guessing the short one only picks a few random spots, while long is... longer – Xen2050 Dec 22 '17 at 1:56
  • The short test changed everything to pre-fail or old-age and reading the documentation now looks like everything's fine. Putting it down to a blip – user1646196 Dec 23 '17 at 11:16

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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