Do hard drive diagnostics tools (i.e. Seatools) ever give false positives?

[PremiumG]

I've been having some issues with my computer BSOD. I ran the Seatools software and it found three hard drives that failed the Long DST (Drive Self Test?).

But since then, I've ran the Long DST on those three drives two more times, and all three have passed.

So my question is are those hard drives back? Would you guys return them?

 

[hoihtah]

I use seatools on a daily basis.
In my experience, seatools does not give false positive.

When you're running these DSTs and seatools detect a bad sector, it tries to move the content on the bad sector to another one. Afterwards, it marks that sector bad and prevents the drive from accessing it.

This is done if it is possible... meaning if the damage is limited to this.

In general however, when it comes to hard drives, I prefer to not have to second guess myself. If you can return it, I would. I'd buy lots of things used and refurbished.... but never hard drives. But that's just me.

 

[corkyg]

Better safe than sorry. However, BSODs are not necessarily the fault of the drives. More often than not it is because of what is on the drives, i.e., the OS, apps, and the Registry.

 

[Binky]

Seagate may not accept the drive without a fight if it passes the self tests. Check their warranty terms. I have no idea if they actually enforce anything (i.e. make the consumer pay for shipping if the drive tests as OK).

 

[sub.mesa]

When you're running these DSTs and seatools detect a bad sector, it tries to move the content on the bad sector to another one. Afterwards, it marks that sector bad and prevents the drive from accessing it.

Yes and it becomes "Reallocated Sector Count" in the SMART data. But if this recovery of a weak sector fails, then you've got a real BAD sector which can NOT be recovered (perhaps spinrite can). In this case, it would show up as Current Pending Sector.

This may also have triggered an error at that point; the utility failed to heal the device and does not want to destroy the unknown data on the bad sector either. Instead, the drive will wait for you to overwrite that sector with data of your own (could be zeroes). If that happens, a 512 byte or 1 sector write to exactly that bad sector, then the drive would write the data to a reserve sector instead and 'map' the bad sector to the new one from the reserve pool. In this case the "Current Pending Sector" in SMART data would be gone and "Reallocated Sector Count" gets one extra again. End result: fixed.

So a possible explanation is that you first had a bad sector, and later you wrote to that bad sector (could be metadata from NTFS that chkdsk fixed) and that lead to fixing the bad sector on disk, which is why Seatools does not give any errors when ran after this event.