Hi there,

maybe here is someone with insight into modern harddrive firmware beyond what's in advertising papers? I need yoiur advice.

It is said that a modern harddrive is capable of replacing sectors found defective.

Recently I dug out a Samsung HD204UI, SMART says it has one defective sector pending (C5=1). From what I know doing a smart erase overwrites the whole disk, and thus should trigger a mechanism to replace the sector with one from the sparer sectors pool.

After performing a full write erase (HDTune Pro) I expected the pending sector cleared, and the "Reallocated Event Count" (C4) and "Reallocated Sector Count" (05) increase by one.

But nothing happens.

Any thoughts why this happens, and what I can do?

Thnx, Armin.

Yes, the hard drive partition may be collapsed due to some reasons, if the issue is minor then the hard drive can repair itself. and the partition tables will be repaired. else try some other methods.

Hi Happie,

thanks for your reply and the information given, but this was not what I was asking for.

Greetings,

Armin.

That's a strange result.

Can you see the bad LBA with smartctl? If so, then try writing to it with a hex editor, eg DMDE.

_________________
A backup a day keeps DR away.

ISTR that at least some Samsung models have a bug in the ATA secure erase command. Perhaps the HD204UI is one of these?

_________________
A backup a day keeps DR away.

Are you really sure, that, when doing a write-erase of a disk, the ATA secure erase command is involved at all?

I thought that the software in use (HDTune Pro or AOMEI Partition Assistant Pro in my case) does simply do write commands with random patterns all over the disk space. Moreover, I read about additional concerns when secure erasing SSDs, but please note that all the disks currently involved are conventional magnetic units.

Nevertheless it may be possible thet the one drive which gives me headaches has indeed a faulty firmware, since I repeated the test with other disks from Samsung which had pending secrors, and with those disks the mechanism worked like expected, but only almost: while the write erase cleared the pending sectors, the reallocated sector count and the reallocation events count remained 0. Either the formerly defective sector was found OK now, or the firmware does "silently" reallocate the sector so the drive can be sold to be fault-free?

I have, over the years, got susipcious. Recently I had, for instance a Western Digital 2TB drive failing, and any 3rd party disk diagnostics tool I tried indicated that the drive was faulty when doing a surface test, but WDs own disk check software did not, marking the disk as "OK". Coincidence or an attempt to avoid warranty returns?

Greetings from Germany,

Armin

Most likely the first option – there would be no point in having two specific SMART fields for sector reallocation operations if those were done “silently”. A sector can be in an inconsistent state which makes it currently unreadable, and then once it's been overwritten it is fully operational again. It can happen, for instance, when there's a power failure during a write operation.


Most likely the second option – the threshold beyond which a drive is identified as “bad” is not the same between the manufacterer's own assessment (usually the 0-100 or 0-200 normalized value has to drop to 0 in the pending sector count and/or reallocated sector count fields) and third-party tools. For instance, when HD Sentinel reports that a drive has a deteriorated “health” status because bad sectors have been detected, it clearly indicates in the “Overview” pannel that “at this point, warranty replacement of the disk is not yet possible, only if the health drops further” (even if by its own assessment the health status has dropped below 70%). But, in practice (at least in my experience), WD's customer service is more lenient than advertised and will replace a HDD with as little as 1 bad sector (by a “refusbished” unit though, which, in my experience, might be a gamble – I once requested a replacement for a 1.5TB drive with 1 supposedly bad sector, which might not even have been physically defective in restrospect, as it was detected by CHKDSK after a sudden power failure, so it might have been recoverable by a simple overwrite, and ended up with a “refurbished” unit which after just a few weeks started to have much more serious corruption issues, so I had to have it replaced again, and the second unit, also “refurbished”, is still operational 7 years later).