Question Quick CMD command to check SSD health

If you see "OK", I believe all that is telling you is that no SMART attribute has fallen below its threshold. That would apply to SATA, but I don't know how a partially degraded NVMe SSD would behave.

I don't think the image in the OP is all that helpful. SMART is not the be-all and end-all of storage drive diagnostics, so just because the SMART status comes back as OK does not mean the drive is fine.

SMART data is potentially helpful, there's no doubt about that, but if you're interested in checking drive health then it's far better to learn a bit about reading SMART stats.

To use the ever-helpful computers-are-like-cars analogy, relying on a single word answer from SMART stats is like waiting until the 'check oil' warning lamp comes on in your car, rather than learning how to check the oil dipstick.

Here's a real-world example - my 'occasional use' laptop has a 512GB NVMe SSD in. Because of a dual-boot experiment that didn't work as well as I hoped, I switched back to Win10 with a clean install. The first few times I attempted it, Windows setup crashed, then from a Windows setup command prompt I ran a full filesystem check and it turned out that the drive has a couple of bad sectors. A couple of bad sectors isn't typically enough to give the user a bad day, but if a bad sector happens in just the right/wrong place, then it will give you a very bad day. What do the SMART stats say? Drive health is good. WMIC says it's OK.

The laptop has run fine since that incident, but if I were to ask advice on this forum about it, I guarantee that some people will say "replace it now". My advice to someone in this position would be depending on their budget, how much they can tolerate downtime and whether they can fix the problem themselves. It's not really a one-word answer, is it?

igor_kavinski said:

Never seen an NVMe drive develop bad sectors. Must've run out of spare sectors. Which brand/model is it?

Click to expand...

I've seen it at least twice with SSDs, and it doesn't require the reserve sectors to be exhausted; the whole point of the reserve is to come into play if a bad sector is found, which is what happened in both cases.

This one is Samsung, I can't remember what brand the previous one was but it came with a Fujitsu laptop.

Here is a case where WMIC is reporting "OK" for a completely dead 970 Evo Plus:

igor_kavinski said:

He didn't post a pic of running that command on his system but if that really happened then I guess this command is only useful for SATA drives?

Click to expand...

It's fine for any SMART-reporting drive but it's only fine up to a very short point.

On my own system, I might look at SMART stats once in a blue moon from a pro-active system care perspective. In doing that, I have learnt enough to spot issues such as:

- bad sectors getting logged (because it's not a certainty that the user will notice weird system activity because this happened)
- CRC errors getting logged (with SATA drives this is usually caused by dodgy SATA data cables, I'd be worried if it happened to NVMe drives because how the hell do I fix that... reseat the drive?)

Out of literally hundreds of failed HDDs, I've only ever seen once "SMART STATUS BAD" HDD and by the time it got to that point it was already unreadable in normal operation.

To diagnose a failing drive, I normally hope for three bits of evidence to align:
1 - Wonky system operation (crashing, freezing-like performance issues are common signs)
2 - Event Viewer > Windows Logs > System: disk/ntfs/ahci warnings/errors citing the suspect drive
3 - Iffy SMART data

If I get all three and say 2 and 3 both report bad blocks/sectors, then it's highly likely that the disk is failing (though I can logically consider the possibility of say a dodgy SATA cable and an otherwise fine drive causing all three if CRC errors are being logged by SMART).

But I've seen HDDs that are only useful as doorstops (in terms of normal operation) have a completely clean bill of health in SMART. I've also seen drives with only one bad sector get replaced because that one occasion caused system downtime which wasn't acceptable to the end user as a potential intermittent ongoing situation, but the fact of the matter is, "SMART STATUS OK". Some programs like CrystalDiskInfo will post their own drive status as 'Caution' in such a situation.

Another example is that the drive itself can detect and deal with bad sectors, remove the sector from normal operation, use a reserve sector instead, all without (as far as the user is concerned) the system skipping a beat. Windows also detects bad sectors and adds the bad sector to its own list rather than the drive's.

One thing that would make that wmic command slightly more useful would be an official Microsoft resource stating exactly under which conditions will make that command report anything other than OK, because at least then one can decide whether Microsoft's opinion aligns with one's own priorities about what constitutes "not OK". It would be even better if it was a configurable setup, so for example I personally would like a warning in the event log if CRC >0 and bad sector > 0 in general.