Never Seen This in my Life...

[sapphirescales]

The weirdest thing just happened today to my personal computer (custom build). I leave my computers on 24/7 so I never shut this computer off. I went to use it today and it was at the BIOS screen saying no bootable devices were available. I went into the BIOS and no devices showed up at all (SSD, hard drive, CD drive, etc.). I tried mucking around with the settings but no matter what I did I couldn't figure it out so I reset the settings to default and the devices showed up again, but I still couldn't boot. I fired up the Windows 2004 installer and when I went to do a fresh install of the OS, the drive showed up as not even having any partitions on it so I did a fresh install of Windows and everything seems normal now. Thankfully I don't keep any data stored on the SSD so nothing was lost, but now I'm afraid to use this computer anymore. It's old (4th gen i7) but I don't want to replace the MB/CPU/RAM if I don't have to.

I have the most recent BIOS update on it (from 2014, unfortunately) and I flashed it again for good measure, but I don't know what the heck happened. The SSD (1TB Samsung 850 EVO) checks out fine. The only thing I can think of is some sort of weird motherboard problem, but I don't know how it would have erased the SSD. Any ideas?

EDIT: Thought I should mention that it might have died when I was encrypting an external drive using VeraCrypt. I started the encryption process before going to bed the previous night. I have no idea whether it completed or not. I checked out the external drive but I think I forgot the freaking password because I can't get it to mount. Either that or the encryption process didn't complete, but it's very possible I forgot the password because it was a new password I've never used before.

 

[Philippe]

Failing SSD. My $0.02...
(Never used the EVO serie -> always Samsung PRO)

 

[sapphirescales]

Failing SSD. My $0.02...
(Never used the EVO serie -> always Samsung PRO)

SMART checks out okay and I ran Samsung's tools on it too. No problems.

 

[seedubya]

I've had this happen twice to me. Exact same symptoms. I have no solution for you - I just replaced the SSD twice. They were both Kingston (forget the model) drives.

 

[sapphirescales]

Well it's been running fine the last few days. All I did was reinstall Windows. I wish I knew for sure it was the SSD, but I can't take chances with this computer. I've ordered a new Ryzen 5 3600 processor and a new ASUS motherboard (with a couple of spares) and 32GB of RAM. When it comes in, I'll probably replace the SSD too just to be safe. This is one of my computers that I use for backing up/recovering client data so reliability is my #1 priority. But now I'm wondering if this might have been a power event that ended up corrupting my drive. My power supply is a Corsair AX1500i so it's top of the line in terms of quality, but I worry about that now too. Maybe I should just replace the whole damn computer.

On a completely unrelated note, my 6TB hard drive has started to fail in another computer of mine (4 uncorrectable sectors) and I'm using this computer to back up the data to an old 8TB drive I have. It was already 95% full so I needed to upgrade it anyway. I just wish I could have used a new drive instead of that old 8TB I had lying around. The problem is, retailers have jacked up hard drive prices since it's Christmas. I just bought a 14TB drive for $189 a month ago and now they're claiming that it was $309 and is now "on sale" for $249. I mean, it's only $60 but I hate being lied to and screwed.

 

[Sky-Knight]

That's a failing SSD, the garbage collection is eating the drive partition table. The disk will eventually simply stop.

And no, you won't get it to fail any sort of actual diagnostic... this is the dark side of SSDs.

 

[Philippe]

ended up corrupting my drive.

Corrupted data on a drive does not prevent it from being detected by the BIOS...
(And btw, SMART is far from giving a perfect picture of the state of a drive, specially a SSD).

 

[sapphirescales]

Corrupted data on a drive does not prevent it from being detected by the BIOS...

That's a failing SSD

But the problem is it wasn't just the SSD that couldn't be detected. Neither the hard drive or CD drive showed up either. This is why I think it's more than just an SSD problem (I'm leaning towards the motherboard).

 

[Sky-Knight]

But the problem is it wasn't just the SSD that couldn't be detected. Neither the hard drive or CD drive showed up either. This is why I think it's more than just an SSD problem (I'm leaning towards the motherboard).

Samsung EVO drives have an internal defense mechanism that disables reads and writes when the device detects faults. I've seen this trip up some mainboards and subsequently lock up the entire SATA bus. So the symptoms to me still say SSD.

Also, as a point of trivia, Samsung Pro devices fault to READ ONLY. Which is a drastic improvement and why I generally recommend getting the Pro drives from that line whenever possible.

Note I've never seen either of these behaviors with WD Blue SSDs... but I don't think they're as well built in this regard either. They'll just do the random data loss thing like most SSDs or defective memory modules and drive you up a wall until you swap it. It's really fun when the device only faults say 3 times a year... often drives will simply persist for 18 to 24 months like this before they finally splatter. Along the way the user just keeps rebooting after hard locks and not complaining too much because again, it only crashes 2-3 times a year.

I'm just glad all of this is mercifully rare!

 

[sapphirescales]

I've seen this trip up some mainboards and subsequently lock up the entire SATA bus.

Do you know if this persists after restart? I had to rest the BIOS settings to default to get it to detect any of my SATA devices (the computer is too old to have m.2).

 

[Sky-Knight]

Do you know if this persists after restart? I had to rest the BIOS settings to default to get it to detect any of my SATA devices (the computer is too old to have m.2).

In one case, but the rest I ran into simply disconnecting the offending SSD and power cycling the unit was enough to get it back.

I don't know if M.2 matters at all in this situation, I haven't had a defective M.2 SSD yet, either NVMe or SATA.

But a ton of modern systems will change the boot order automatically when they detect a new boot device. These systems can wind up with a corrupted boot order if a device is defective when it's scanned... that will need a CMOS clear to flush out.

 

[HCHTech]

Samsung EVO drives have an internal defense mechanism that disables reads and writes when the device detects faults. I've seen this trip up some mainboards and subsequently lock up the entire SATA bus. So the symptoms to me still say SSD.

Also, as a point of trivia, Samsung Pro devices fault to READ ONLY. Which is a drastic improvement and why I generally recommend getting the Pro drives from that line whenever possible.

Wow, thanks for this. We standardized on Samsung a few years ago and have installed hundreds of drives, usually non-pros for consumer and pros for commercial use, but I didn't know this bit. In all of that, we've only had 2 failures, one non-pro was DOA, and the other one was acting squirrelly enough that we replaced it on a hunch to prove it was the drive. Samsung replaced it without complaint, but the new drive we got as a replacement was DOA! We're in the middle of claiming that drive now - ha.

 

[Sky-Knight]

Wow, thanks for this. We standardized on Samsung a few years ago and have installed hundreds of drives, usually non-pros for consumer and pros for commercial use, but I didn't know this bit. In all of that, we've only had 2 failures, one non-pro was DOA, and the other one was acting squirrelly enough that we replaced it on a hunch to prove it was the drive. Samsung replaced it without complaint, but the new drive we got as a replacement was DOA! We're in the middle of claiming that drive now - ha.

I love Samsung drives, but yeah that little quirk of firmware can really get you in the right circumstances. I've only had 1 EVO fail, and it was thanks to poor firmware garbage collection. Firmware updates have long since fixed this, and even this singular case had a happy ending because the fix was just to power the drive without a data connection for a few hours so garbage collection could catch up.

That drive "died", because Windows update on Windows 7 went nuts... When the unit came in, I fixed it... upgraded it to Windows 10... and sent the entire mess back. Three years later the client in question fired me, the unit was still in service.

So I mention the above as a matter of extreme caution, but given how many Samsung's I've used myself... it's not like I'm terribly worried about it. It's just that for high end use, you'll probably want the Pros. I know I use Pros at least in servers, of any kind... even the overgrown desktop kind.

I've sold hundreds of Samsung EVO drives in Nexgen Appliances too, never had an issue there. But, those are Untangle servers, with presumably admins backing up configurations that make the storage essentially disposable. Recovery really isn't reasonable in those circumstances! The thing quits? You replace it and restore your backup. Don't have one? Bad admin! No cookie!

 

[sapphirescales]

@Sky-Knight Okay I'll take this under advisement. I never knew any of this and I've probably sold 2,000+ Samsung SSD's over the years. I've seen a few dead ones (all but one related to firmware issues) but this has never happened to me. I think I'll take this SSD out of service and stress test the sucker for a couple of days/weeks to see if I can get it to either fail outright or throw a SMART error of some kind.

 

[Sky-Knight]

@Sky-Knight Okay I'll take this under advisement. I never knew any of this and I've probably sold 2,000+ Samsung SSD's over the years. I've seen a few dead ones (all but one related to firmware issues) but this has never happened to me. I think I'll take this SSD out of service and stress test the sucker for a couple of days/weeks to see if I can get it to either fail outright or throw a SMART error of some kind.

It likely won't, one day it'll simply stop. And until then it'll pretend it's happy.

But yes, this is rare for "good" SSDs, any of them with real names behind them. Samsung in particular has an amazing track record, so it doesn't shock me that you've never seen this behavior. I'm more likely to see it I assume due to the environment here in Arizona. We have power problems here... ugly ones. They make electronics go stupid.

But were I you, I'd not change anything but that drive. The odds of you finding another one are rather slim.

 

[sapphirescales]

@Sky-Knight Okay so I'm using this computer to back up the data from a failing 6TB hard drive. It was running for 3 days and then Unstoppable Copier (the program I'm using to back up the drive) just closed by itself when the copy process was at about 59%. The computer did NOT restart (it still shows an uptime of 3 days in Task Manager), but the program crashed. Could this be caused by a failing SSD? I've been using Unstoppable Copier since 2010 and it has NEVER crashed on me (it's a very small program). My theory is that the program crashed because data that was stored in the paging file (on the failing SSD) became corrupted or wasn't stored correctly.

If I look under Resource Monitor, the number of hard faults is consistently between 80 - 100, which is a lot more than normal. I'm only using 15% of available memory. I have no idea if this is normal behavior when using Unstoppable Copier.

 

[Sky-Knight]

@sapphirescales You know all the strange niggly problems you can get if you've got a defective RAM module?

In short all of those nightmare problems that are ghosts in the machine you can spend weeks chasing down?

Each and every stupid one of them could be a bad SSD. Because it's RAM, and it fails like RAM. But unlike RAM, you can't just run memtest on it and have it find it. Well maybe you can... but I've never found a tool that tests an SSD like RAM.