In the past month I experienced three hard drives suddenly failing for seemingly no reason:
- A 500GB SATA Samsung hard drive in an RF511 notebook (6 Series/C200 Series Chipset Family 6 port SATA AHCI Controller), after 1.5 years of normal usage.
- A 160GB ATA WD 1600JB hard drive in a desktop computer (GA-K8NF9 motherboard) after over 5 years of usage.
- A 500GB SATA TOSHIBA MQ01ABD0 hard drive in the same notebook as in 1 (warranty replacement), after 113 hours (!) of power on time. SMART dump after the failure: http://slexy.org/view/s2p1eNuA1e
All of the failures happened roughly at the same time (a few weeks apart) and in the same room (in an ordinary house). Both computers were running Linux (they had different kernels and drivers though). Other than that, I can't point out any similarities in the failures.
The scenario is as follows:
- The computer is functioning properly without any signs of upcoming failure.
- Suddenly, the disk fails a lot of read operations (Linux reports failed READ FPDMA QUEUED commands). Off-line SMART self-tests fail as well (read failure).
- Trying to write to the sectors results in even more damaged sectors. I managed to get hundreds of bad sectors in a few hours on a disk that was healthy the day before.
Things I've checked so far:
- Replace the SATA cable - it didn't help.
- Plug the damaged disk into another computer - the disk appears just as damaged.
- Replace the disk. In the notebook, the new disk failed after two weeks. In the desktop computer, the disk is still functioning properly.
I have no idea at all what could be causing it. All I can suspect are environmental causes of some sort.
