HDD bad sectors – chkdsk versus hdtune

bad-sectorschkdskhard drive

I bought new HDD WD Blue (WDC WD10EZEX) as a secondary storage media (no RAID or anything, just a bunch of big video files on NTFS). There were a few problems at the beginning, Windows freezed at startup, a few chkdsk sessions were automatically invoked by the system. I have changed cables and problem was gone for a week or two.

However:

  • Chkdsk now reports 4 kb in bad sectors.
  • HD Tune error scan shows no problems and tool reports Health Status as "OK".

How should I interpret such situation?

Best Answer

  • Once a bad sector has been marked as such, it's replaced by a sector from the spare area, transparently to the operating system, which means that a subsequent scan with a tool like HDTune will show no problem (if no new bad sector has developped in the mean time).
  • What CHKDSK considers as a bad sector can sometimes be in fact a magnetical defect/inconsistency (as opposed to an actual hardware malfunction) which may have be fixed by a write operation (or a so-called “low-level formatting”, which actually means filling the whole surface with zeroes), or a tool like HDD Regenerator / SpinRite. I once had a 2TB HDD (WD20EARX) with a problematic area, that caused CHKDSK to fail repeatedly with a “not enough space” error, even after I had deleted more than half of the content. I ran HDD Regenerator, which fixed the defective/inconsistent area (apparently it was really fixed, rather than replaced), CHKDSK could still not finish (but WD's Data Lifeguard Diagnostics “quick test”, which failed before the HDD Regenerator scan, was now normal), then I ran a defragmentation, then CHKDSK could proceed through the end, further fixing the file system, and the HDD is still in use 2 years later (with 0 sector reallocation according to SMART data). (Data Lifeguard Diagnostics also has a “long test”, which I tried at the time : it reported “bad blocks” and proposed to fix them, with a warning of potential data loss, and no exact information as to where those bad blocks were found ; it may have fixed the problem as well but I did not run it through the end, looking for a more “verbose” alternative in an effort to understand what was going on, rather than trusting a “black box” to fix the issue. With HDD Regenerator I knew exactly which sector was defective, which then allows to check the corresponding file for potential corruption, using WinHex or NFI to find the correspondance.)
  • Still, it's not normal to have that kind of issue with a brand new HDD, and unless the cause is clearly identified (possibly bad cables in this case), I would probably send it back.
  • The problem when you send back a HDD in RMA to get another one is that most of the time you get a “re-certified” one, i.e. one that was in use, had a problem, and was fixed by the manufacturer (at least this is the case with Western Digital, I never had to return a Seagate HDD). It's supposed to meet the same standards of reliability, but it's more like a lottery, and even if it does work fine it can be in quite a poor state aesthetically, and then you can't complain as long as it's functioning and passes the manufacturer's fitness test -- I once received a “re-certified” Western Digital HDD which still works fine but the metal cover is scratched on the sides. I once had a single bad sector on a HDD which I knew was caused by a power failure during a write operation, and sent it back to RMA, but then regretted it : the replacement unit became defective soon afterwards, developping hundreds of bad sectors, and corrupted some files, whereas the original one might have worked fine once that bad sector had been identified and replaced (it may even have been fixable by HDD Regenerator, which I didn't know about then).
Related Question