• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

SSD cannot remove partitions

Joined
Mar 11, 2009
Messages
1,778 (0.30/day)
Location
Little Rock, AR
System Name Gamer
Processor AMD Ryzen 3700x
Motherboard AsRock B550 Phantom Gaming ITX/AX
Memory 32GB
Video Card(s) ASRock Radeon RX 6800 XT Phantom Gaming D
Case Phanteks Eclipse P200A D-RGB
Power Supply 800w CM
Mouse Corsair M65 Pro
Software Windows 10 Pro
Hello everybody.

So, I've got this SSD. It's a Samsung MZ-7TE2560 (OEM from a laptop). I had it in the laptop, had windows 10 running on it just fine for several months. One day it just won't boot up. It had the "BCD corrupted" windows boot error. So I reinstalled windows 10, no biggie. Went on about my life, this was about 2 months ago.

The other day, I get the same error again. I suspect that hibernation is somehow to blame, but that isn't the focus of this thread.

I couldn't reinstall windows, because I couldn't delete partitions on it, windows installer wouldn't let me. So I boot up windows rescue tools and attempt to use diskpart clean.

So now here's where it gets weird.

Diskpart does its thing, and I list part on the SSD, no partitions, we're golden, right? Wrong. I boot back into the windows installer, and as if by magic, the partitions are back....

I do it again. I do diskpart clean all. And again. And again. I make sure the drive isn't read only. I google my butt off. Nothing. I pull the drive out, plug it into a usb dock, and use diskpart clean all on my other computer. No dice. The partitions keep reappearing. I use datalifeguard diagnostics to rewrite the drive with zeroes. Nothing is seeming to work here.

How in the holy terror are these partitions still there? It tells me of course by now that they're corrupted and can't be formatted. The only info I can find on this subject is how to delete the partitions in the first place, no info on what happens if they DON'T go away.

Anybody have any experience with this?

Thanks in advance.
 
boot to a live cd linux and run DD. should fix ya.
 
It sounds to me like the drive might have gone into Read-only mode. Which is a fail-safe mode when the drive has run out of extra sectors to re-allocate.

Have you checked the SMART data?
 
boot to a live cd linux and run DD. should fix ya.

That was probably going to be my next step, though it's frustrating that windows tools can't do it.

It sounds to me like the drive might have gone into Read-only mode. Which is a fail-safe mode when the drive has run out of extra sectors to re-allocate.

Have you checked the SMART data?
I was still in the process of getting to the bottom of the SMART data when I posted this. Samsung's tools won't recognize the drive since it's an OEM drive, and every other tool I could find had "unknown" values, so they were useless. Finally found crystaldisk that could read them all. Says wear leveling count is bad. Current: 1 Worst: 1, threshhold: 5. That doesn't make any sense....
 
That was probably going to be my next step, though it's frustrating that windows tools can't do it.


I was still in the process of getting to the bottom of the SMART data when I posted this. Samsung's tools won't recognize the drive since it's an OEM drive, and every other tool I could find had "unknown" values, so they were useless. Finally found crystaldisk that could read them all. Says wear leveling count is bad. Current: 1 Worst: 1, threshhold: 5. That doesn't make any sense....

your reading the values wrong. does crystal disk say the drive is bad? if so @newtekie1 is right.
 
your reading the values wrong. does crystal disk say the drive is bad? if so @newtekie1 is right.

Yep. That's what I mean, it doesn't make sense. It says wear leveling check is bad, but the counts are as I said above.

So if that's the case, how do I get it out of this read only mode? Surely this drive is not *actually* gone bad.
 
Yep. That's what I mean, it doesn't make sense. It says wear leveling check is bad, but the counts are as I said above.

no your understanding of the counts are wrong. the drive is bad.

The meaning varies by manufacturer but I am going to take it that in your drives case 5% health REMAINING is the thresh hold and the drive is currently at 1% aka DEAD
 
no your understanding of the counts are wrong. the drive is bad.

The meaning varies by manufacturer but I am going to take it that in your drives case 5% health REMAINING is the thresh hold and the drive is currently at 1% aka DEAD

I see. I have a sandisk that reads opposite of this one then. I read some more on the read only mode, I guess it goes into that to make sure that data is still available, but no more write cycles left? That's insane, the laptop is only about 2 years old. Surely it can't be worn out already?
 
I see. I have a sandisk that reads opposite of this one then. I read some more on the read only mode, I guess it goes into that to make sure that data is still available, but no more write cycles left? That's insane, the laptop is only about 2 years old. Surely it can't be worn out already?

it really depends. SMART tools can only read the data in the drive registers. Its upto the manufacturer to program what they mean. which means in some cases we have to kind of guess. As for it dying its not out of this world. it really depends on what happened to the drive. IDE mode? no TRIM? controller failure? unfortunately finding those answers is much harder.
 
IDE mode? no TRIM? controller failure? unfortunately finding those answers is much harder.
Also the firmware the controller runs could of been responsible for drive going into read only mode.
 
hiberfil.sys (6GB+) is Hibernates restore "cache", it writes to SSDs almost as much as Superfetch or Indexing does. Being in a laptop Hibernate is used a lot more often than on a desktop, so its no surprise the SSD wore itself out after 2 years.
 
If diskpart didn't fix it. (if you did it right) than the only hope it Samsung or their software and secure erase. You can also check the drive status

The drive is not out of writes in 2 years. Not possible. But it may be damaged
 
You might try the SSD LIfe trial to see how much has actually been written to the drive.

hiberfil.sys (6GB+) is Hibernates restore "cache", it writes to SSDs almost as much as Superfetch or Indexing does. Being in a laptop Hibernate is used a lot more often than on a desktop, so its no surprise the SSD wore itself out after 2 years.

Yep, hibernate is a massive writer on drives. Especially with Windows 8/10 that doesn't shutdown but instead hibernates every time you tell it to shutdown(unless you change the setting to make it properly shutdown). Plus Hybrid-Sleep means that when the computer goes to sleep, it write to the hiberfil.sys like it is hibernating too, so if the power goes dead when in sleep, you can still wake it from hibernation.

Moral of the story: Disable hibernation if you have an SSD.
 
I say disable hibernate even if you don't. Shutdown when you need to, boot times are much faster than the xp days of old. granted still nowhere near DOS days or in some cases win 3.1 days. But most people are under 15 second boot these days, so hibernate becomes silly. Also modern rigs power states are so much more efficient that these days I just turn off the monitor if I don't want to shutdown. Sleep and Hibernate aren't as useful as they once were.
 
Thanks for all the insight. I had no idea hibernation was such a drain on an SSD. That being said, I usually did shut it down properly (or at least I thought I did, I didn't realize that Win10 actually hibernates when you shut it down...) The two times that I got the windows BCD error happened after a long hibernation spell so... I dunno.
I've already disabled hibernation after putting in a new SSD and reinstalling windows 10 (didn't want to wait until the other SSD got fixed or not) so that's good.

Unfortunately, as it's an OEM drive, the samsung software doesn't recognize it, so I can't use samsung's software for anything.
 
I say disable hibernate even if you don't. Shutdown when you need to, boot times are much faster than the xp days of old. granted still nowhere near DOS days or in some cases win 3.1 days. But most people are under 15 second boot these days, so hibernate becomes silly. Also modern rigs power states are so much more efficient that these days I just turn off the monitor if I don't want to shutdown. Sleep and Hibernate aren't as useful as they once were.

I agree with Hibernation, but I still use sleep. The wake-up is instant, a lot quicker than a full startup even with an SSD. But Hibernation is pretty useless these days, it isn't really any faster than a normal startup. There is a pretty significant power saving with sleep over just letting the computer sit idle though.

Thanks for all the insight. I had no idea hibernation was such a drain on an SSD. That being said, I usually did shut it down properly (or at least I thought I did, I didn't realize that Win10 actually hibernates when you shut it down...) The two times that I got the windows BCD error happened after a long hibernation spell so... I dunno.
I've already disabled hibernation after putting in a new SSD and reinstalling windows 10 (didn't want to wait until the other SSD got fixed or not) so that's good.

Unfortunately, as it's an OEM drive, the samsung software doesn't recognize it, so I can't use samsung's software for anything.

Make sure to disable "Fast Startup" and not just disable hibernation through the power control panel, but actually disable it through the command prompt using the powercfg.exe /hibernate off command.
 
Now I'm confused....

(Yes I verified this is the correct disk, the computer I'm running this from does have an identical drive in it)

This says it's good.

disk.png
 
According to SMART the drive has been overwritten 5115 times (13FB in Wear levelling count), supposing it has the same value as my Samsung 850 EVO, which it increases everytime the entire SSD capacity has been overwritten.

5115*256GB = 1309440 GB writen :eek:
Anyway, most of the values indicated are way off, for example it says the SSD has lost power (POR Recovery Count) 15456 times.
 
Last edited:
I had this issue a while back with two other samsung SSD's. I used linux gparted to delete all of the partitions then plugged them in USB onto a windows machine and formatted them in disk manager. They work good as new now.
That's what I'm going to do, haven't had time to set it up. I used to keep a linux mint live USB laying around for just such an occasion, but I've lost it.

EDIT: misread that, I thought you were just talking about using DD on linux. Didn't see the Gparted. I'll try that out.
 
wait. when its plugged in have you tried using diskpart from the cmd?

I do it again. I do diskpart clean all.

nvm i didnt take my pills today.
 
Last edited:
Yep. that's what was so weird. Clean and clean /all seemed to work fine. Said they were successful.



Thanks, I'll try that too, if needed. :)

Parted Magic>Enhanced Secure Erase.
 
I agree with Hibernation, but I still use sleep. The wake-up is instant, a lot quicker than a full startup even with an SSD. But Hibernation is pretty useless these days, it isn't really any faster than a normal startup. There is a pretty significant power saving with sleep over just letting the computer sit idle though.



Make sure to disable "Fast Startup" and not just disable hibernation through the power control panel, but actually disable it through the command prompt using the powercfg.exe /hibernate off command.
added that command line to my sig for this reason. :rolleyes:


wait. when its plugged in have you tried using diskpart from the cmd?



nvm i didnt take my pills today.
a bong is holds more pills




Appears its only for HDD not SSD
No guarantee of data removal (e.g. DBAN does not detect or securely erase SSDs)
 
Last edited:
Back
Top