Question Help! SSD Raid0 not bootable but everything is OK on it!

[crowley_dot]

Hi guys,

I have a MSI GT72S 6QE laptop where the OS is installed on a PCIE RAID0 SSD (I know, Raid0 sucks, I won't make that mistake again when buying a laptop).
On of the two SSD's in the RAID kept getting not recognized at all, the room temperature seemed to be the reason (I know how it sounds but everytime my room got under 20°C the SSD wasn't recognized) but it also happened seemingly at random.
I then would have to reboot my computer a number of times for the SSD to be found and then everything would work absolutely fine.

I got tired of it so I opened the laptop and reseated the SSD's in the RAID structure and booted my computer.
One of the disks was marked as "Error Occured" in the "Intel Rapid Storage" section of my BIOS but besides that every test I could find told me the RAID was OK.
I installed another windows on an HDD that was also in the computer so I could mark the drive as normal in "Intel Optane".

Here is the drive that was marked as error:

Here is the other :

Here is the raid, and when I do "Verify", it passes 100%:

Now in the BIOS both SSDs are marked as "RAID member" and the RAID is in status OK and Bootable "Yes".
After that I did every volume test available, TestDisk can see EVERY file in the Windows partition and the boot manager still sees a Windows boot on that volume but when I try to boot on that Windows, I just get a black screen.


Here is the disk in TestDisk:

What's in the RAID is all OK, Windows is in "OS_Install", obviously

And here are the contents of OS_Install:

Chkdsk finds no error at all. TestDisk doesn't see any lost partition and tells me the boot record and backup boot record are fine and yet I can't neither boot on that windows or repair it, when I try the "Startup Repair" the laptop reboots and MSI tells me "checking drive for errors, this could take hours" but stops after a few minutes and reboots.
If I try to reinstall windows from the Cloud ou local image it says right away that there is a problem without saying more and goes to the previous screen.
I'm at my wit's end. If the SSD was really dead, could I do everything I did above? I don't think so but I'm not a pro.

In the other windows installed on the HDD, the drive appears as the letter "D" and isn't accessible, it tells me "access denied" if I try to go in, but TestDisk can with no problem and chkdsk can find and test every volume in the RAID.

What is maddening is that both SSD's seem absolutely fine, the RAID too, Optane, testdisk, chkdsk, etc. find no error and the boot manager sees my windows install on the RAID but cannot boot or repair it, I'm at a total loss.
I can't believe that the SSD is dead and yet I can acess and copy every file on the volume with TestDisk, but I don't know what to do next.

I need help guys, I feel like it's a little stupid thing like remaking the boot record or something (but it's EBT, not MBR, and I don't know much about either) like that and I would be able to boot on it no problem.
I really don't want to restart on a fresh windows but if I have to I will, but I don't know how to do even that, the RAID seems both fine and failed at the same time.

 

[Ralston18]

Windows OS - edition/version?

Redact your "Test Disk" screens. You are revealing personal identifiable information.

 

[crowley_dot]

Sorry it's Windows 10 Family.
What information on the testDisk screens? I don't really care that ppl see what's in my C:

 

[Deleted member 14196]

Your email is listed

 

[crowley_dot]

Chris Grenier is the guy that made testdisk, not me.

 

[USAFRet]

1. Why the RAID 0?
2. Was/is there a known recoverable full drive backup of this?

 

[crowley_dot]

1. Not my choice, came with the laptop like that, but I wont't ever buy another one like that!
2. Not recent but a few months after buying the computer I saved the recovery ISOs with MSI Burn Recovery on an USB hard drive. But it's from 2016.

 

[crowley_dot]

The problem seems to be in the "OS_Install" partition and yet I can copy the files from it, I'm in the process of saving my entire Users folder for exemple, and everything is copying fine, that's what's really pissing me off is that no errors scan can find anything wrong, I can access the files with testdisk no problem but cannot boot on it...
If I could just repair what ever the hell is wrong on the partition and the windows install on it. Damn...

 

[USAFRet]

The problem seems to be in the "OS_Install" partition and yet I can copy the files from it, I'm in the process of saving my entire Users folder for exemple, and everything is copying fine, that's what's really pissing me off is that no errors scan can find anything wrong, I can access the files with testdisk no problem but cannot boot on it...
If I could just repair what ever the hell is wrong on the partition and the windows install on it. Damn...

This would the point in time where you do a full OS reinstall, on a single drive. No RAID 0.
Actually, replace one of those small 128GB things with a larger drive. 500GB or so, then do an OS install on that.

 

[crowley_dot]

Yes but I'm trying to avoid that.
If I'm going to have to restart from scratch I'll do it with a new computer, not a 5 year old one, so if I can't save this one as is then I'll just copy everything I can from it and start on the new laptop (probably an MSI Pulse GL76, without a RAID0).
I really think there has to be a way to repair that install and boot on it, all those tests wouldn't be finding the drives perfect like that if it was dead; I'm hoping like hell anyway.

 

[fzabkar]

Could it be that the faulty SSD has reverted to readonly mode?

 

[USAFRet]

Yes but I'm trying to avoid that.
If I'm going to have to restart from scratch I'll do it with a new computer, not a 5 year old one, so if I can't save this one as is then I'll just copy everything I can from it and start on the new laptop (probably an MSI Pulse GL76, without a RAID0).
I really think there has to be a way to repair that install and boot on it, all those tests wouldn't be finding the drives perfect like that if it was dead; I'm hoping like hell anyway.

I completely get that you don't want to.

But this is one of the problems with a RAID 0.
And especially without a full drive backup that you can recover.

 

[crowley_dot]

Could it be that the faulty SSD has reverted to readonly mode?

I'm asking myself the same question but can't find anything about it in the BIOS or Optane, everything is marked as ok.
Are there any app out there that can check individual SSD's health in the RAID instead of the whole partition? Only Optane "sees" it as two disks.

 

[USAFRet]

Are there any app out there that can check individual SSD's health in the RAID instead of the whole partition? Only Optane "sees" it as two disks.

If the system fails to run, your investigation options and tools are severally limited.

This is why a RAID 0 should never ever be run without known good full drive backups.

 

[fzabkar]

Smartmontools can retrieve the SMART data from individual drives behind some supported RAID controllers, but I have no idea about Optane.

 

[crowley_dot]

If the system fails to run, your investigation options and tools are severally limited.

This is why a RAID 0 should never ever be run without known good full drive backups.

Well I can't boot on the Windows installed on the OS_Install partition on the RAID but I installed another Windows 10 on the HDD that's also in the laptop, so I don't have to use only bootable tools.

Smartmontools can retrieve the SMART data from individual drives behind some supported RAID controllers, but I have no idea about Optane.

I have smartmontools on a bootable UBCD or other USB stick but can't boot on it idk why, must be too old it was an UBCD I made for work a while ago. Does smartmontools exists for Windows?

 

[crowley_dot]

UPDATE:

I've tried to fix the windows install with sfc /scannow with offbootdir=D:\ and offwindir=D:\Windows\ from the Windows Boot USB command prompt, it fixed errors successfully but I still get a black screen when trying to boot on that Windows.

So I tried DISM to fix the install, with a DISM /Image: D:\offline //Cleanup-Image /Restorehealth with the install.esd from the USB as source but I get an error 2 (or 5 depending on the command) telling me "Unable to access the image", so then again the drive seems OK but can't be accessed by "elevated" tools, but "low level" tools can see every file and doesn't find any errors...
It's maddening seriously.

Is there a disk attribute (diskpart doesen't see it as readonly, I checked) or something that I missed?

PS:
When I first opened the laptop I didn't note the order of the SSD's in the RAID "structure", and I switched them from their ports several time, so I don't remember at all in which order (there are only two ports where the disk can fit on that "structure") they where.
If I have put them back in the wrong order, could I still access the files on the partition?

Furthermore, from the command prompt when I boot on the windows USB, I can do dir d:\windows and it shows me the files, but if I do dir D:\ it tells me "access denied", does it say anything to you guys?

 

[USAFRet]

Furthermore, from the command prompt when I boot on the windows USB, I can do dir d:\windows and it shows me the files, but if I do dir D:\ it tells me "access denied", does it say anything to you guys?

It indicates to me that you should try to locate and copy any personal critical files to some other storage location.
Then....full wipe and reinstall, without this RAID 0 config.

That is what I would do...

 

[crowley_dot]

Yeah I got it, I'm trying not to, I've already saved the files from it with testdisk but now I'm trying to save the OS install. I didn't even have many files to save, I never put anything important in the OS partition, so except for my Firefox profile and a few odds and ends in the User folder there isn't anything really important on it, except for the windows install.
If I can't then forget it, I'm already buying a new laptop (without a RAID0, I didn't know what it was when I bought the dominator), but I wanted the switch to be as painless as possible; so with the older computer working on the side that would be great.
But I'm not going through a reinstall and new setup of all my preferences and software just to do it again in a week when I get my new laptop anyways.
If I can't save the OS I'll just bin the RAID, format the 1To HDD and install a clean windows on it then gift the laptop to my nephew, it runs pretty slow on the HDD but it's better than nothing.

 

[crowley_dot]

I just used MiniTools Partition Wizard on the other Windows installed on the HDD, it sees the RAID perfectly, sees all the partitions OK like testdisk does. I did a surface test on the whole disk and on individual partitions, it passed with flying colors. Re-did a chkdsk on the OS_Install partition just in case but 0 errors again. I can explore the partition also with MiniTools and I can see everything where it's supposed to be. TestDisk tells me the boot record and mft are both OK and perfectly matched to their backup/mirror.


DiskGenius sees both TOSHIBA SSD's OK, the SMART info is all marked Good and when I compare all the infos on both disk they seem to match except for the "Total Host Read/Total Host Writes" values, is that normal (I'm presuming it's the amount of data written on each disk right?) I did a bad track search that gave me all Excellent, Good or Normal tracks/cylinders. (Most of them Excellent, around 10K good and 3 Normal, no bad tracks found). Again it also sees every file in the main partition and can even explore the SYSTEM and WinRe partitions and they look OK to me.


That's one hell of a headscratcher! If one of the SSD's or the RAID are really damaged then could I do all this? But then if they're OK, what the hell is locking the main partition like that?

UPDATE:
Hey something I just noticed: in DiskGenius it shows the Windows I installed on the HDD as GPT, but it didn't create a SYSTEM or MSR partition on the SATA HDD, it's using the ones on the RAID to boot right? So at least those partitions are accessible, why then the main partition isn't?
I want to try to pull the HDD from the laptop and see what happens when I boot, could it have mixed the main partition and try to boot on the wrong one even when I chose the SSD in the boot manager?