1

I am really baffled that one of my movie files doesn't seem to like Samsung HDDs and I was wondering if certain data patterns can cause or at least favor data corruption over time.

It first started when I tried to read back some MKVs from an external 3.5" Samsung HDD. All files copied fine, except this particular file that caused the drive to slow down and click every few seconds. After about two minutes, Windows gave up copying telling me that the source file could not be read. I assumed a corrupt drive and viewed the SMART parameters to find that pending sector count was no longer 0 and offline uncorrectable error count was high. The extended SMART self-test was aborted by the drive telling me that a surface error occurred. I had the same data stored on an identical Samsung HDD and on a WD portable HDD. I tried to copy from the second Samsung HDD and the same file caused errors. Exactly the same symptoms as with the first Samsung HDD. Clicking noise, aborted copy, current pending sector count was 4, high offline uncorrectable error count. Fortunately, I was able to copy the file from the WD drive without trouble. Having two broken Samsung HDDs, I wanted to throw them away, but before doing so, wipe them to be sure all data is gone. I used DBAN to erase the disk surface with zeros and no write errors were reported. I checked the SMART status of the disks and current pending sector count was back at 0, offline uncorrectable remained the same, but after cycling the power, it was at 0. I used a data integrity tool called H2testw to check the disks and even after multiple runs, the written data (occupying the whole disk size) was read back fine and without errors. I then copied the MKVs to the Samsung HDDs and wanted to copy back the files to check if everything went well, but then Windows aborted the copy back from the Samsung disks at the same file that was causing trouble the first time.

So, is this all just a big coincidence or how come one file can cause a drive to fail?

2
  • 1
    Just a coincidence. File content can't cause corruption. You could check the product packaging and see if it says something like, "Will store any file except a particular one of smares' movies", in which case, I'll retract my comment. :-) – fixer1234 Feb 29 '16 at 21:45
  • No; HDDs are designed to be partitioned and hold data – Ramhound Feb 29 '16 at 22:17
1

Just a coincidence. Once the drives remapped the bad sectors, they were no longer accessed when doing full writes/scans of the disk (Is the total number of reallocated sectors still the same for both drives?).

Also, drives tend to fail in batches. If you bought both samsung drives at the same time/same store, there's a high likelihood that they came from the same lot of drives and share any manufacturing defects that affect that lot. (This is why you should change up brands or distributers when building RAID arrays - That way when one drive fails, you reduce the risk of losing a second to the same defect during the rebuild)

2
  • If the drive remapped, shouldn't reallocated sector count be non-zero? It is my understanding that a drive, should it encounter read errors, marks the sector as possibly broken and increases current pending sector count. If repeated reads of that sector fail, the sector is remapped, current pending sector count is decreased and reallocated sector count is increased. In my case however, after the full wipe, both reallocated and pending sector count were 0. – user410854 Mar 3 '16 at 23:14
  • Yes, that is correct. I would assume you don't have remapped sectors if that's the case. – Darth Android Mar 3 '16 at 23:22

Your Answer

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