RAID0 - Abnormal Status Reported by Rapid Storage Technology UEFI Driver

[Jaro]

Hello,

I have the RAID0 built on 2 NVME disks. This RAID is for data only and does not contain any bootable partition.

It worked fine till the upgrade of my Intel Core CPU from 12th to 14th generation.

Afterwards, my system stopped booting due to blocked access to all NVME disks.

I had to disable the VMD setting in the BIOS and then updated the Intel RST to 20.1.0.5817 RSTVMD Driver. Then I enabled again VMD in the BIOS.

Right now Windows 11 system which is on a separate disk (none RAID), works fine.

But my partition for data which is hosted on the RAID0 volume is broken.

During booting the BIOS reports Abnormal Status Reported by Rapid Storage Technology UEFI Driver. In Windows, the Rapid Storage Technology application, in the “Manage” tab, reports missing one of my disks.

Both used Samsung 990 Pro NVME 4 TB disks are fine and are detected in the BIOS properly.

Unfortunately in the Intel RST application, one of the mounted disks is reported as unavailable.

But this missing disk is visible as a non-RAID member next to the RAID.

The problem is that I do not have the backup. I tried to restore my RAID0 to live, but the management application does not provide any option for that.

How do I force attaching my second NVME disk again to the broken RAID0 without losing the data of course?

I am pretty sure that the disks are untouched. Can I edit the RAID metadata somewhere to manually fix the problem?

Best regards,

Jaro

[RobbieR_Intel]

POSTBACK:

Hello Jaro,

Thank you for providing such detailed information about your setup and the steps you have taken so far. I understand how critical this data is to you and how frustrating it must be to face issues after upgrading your system. Your detailed observations abou the RAID0 behavior and the disk's status are invaluable in troubleshooting the problem.

Please answer the following questions:

Are there any SMART Errors?

When you initially set up the RAID0 array, did you use Intel RST for the motherboard's RAID configuration utility?

Are there any specific settings or metadata visible for the broken RAID in the Intel RST app?

Have you attempted any RAID recovery tools or software?

Please generate an SSU report to help me further analyze important details on your system. To generate the SSU report, please refer to the article: How to get the Intel® System Support Utility Logs on Windows*. Please send us the generated SSU.txt file.

I look forward to your response!

Best Regards,

Robbie R.

Intel Customer Support Technician

[RobbieR_Intel]

Hello Jaro,

I wanted to check if you had the chance to review the questions I posted. Please let me know at your earliest convenience so that we can determine the best course of action to resolve this matter. 

Best regards,

Robbie R.

Intel Customer Support Technician

[RobbieR_Intel]

Hello Jaro, 

I have not heard back from you so I will close this inquiry now. If you need further assistance, please submit a new question as this thread will no longer be monitored.

Best regards,

Robbie R.

Intel Customer Support Technician

[Letni_sux_big_time]

Here's the solution. Firstly, intel based bios and drivers for VMD are seriously unreliable. If you have a RAID setup, it will get corrupted during Bios updates or in some motherboards even if you turn off VMD and turn it back on!! I really need nvme raid 0 on my setup so after losing my data for the thrid time, i found out a way to recover the data from a failed RAID0 array. This works for other RAIDs too.

1) Make sure you have the same problem. After a bios update or whatever, the RAID definition in your motherboard's VMD has failed. Only 1 or none of the drives are detected as RAID. HOWEVER, the drives are detected as separate non-raid entities.
2) Disable VMD since the RAID metadata is gone for good. (Thanks to Intel!!) At this point ur drives and data are fine so we'll attempt recovery at software level. Make sure drives are configured as normal nvme or ahci/ide and not RAID.
3) Download free third party software called Reclaime RAID recovery.
4) Select the affected drives from the list and select the appropriate RAID type.
5) The software will delect the RAID at a logical level and will give the output of the analysis. There are 2 paths from here.
a) Easy way (but paid): You can continue with recovery in the same Reclaime software by purchasing the recovery license.
b) Free way: In the Reclaime analysis result, it will allow you to logically read the RAID volume and dump it as a windows vhdx virtual hard drive format for free. Download a free open source utility called TestDisk. Go to This PC->Manage->Storage management ->Actions->Mount/Attach VHD. After the VHD is mounted, it will show up as an allocated drive without a drive letter.
6) Run TestDisk and analyze this virtual drive. It should be able to let you browse through your RAID volume files as it was before it got messed up. Select files you want to copy and save it in a different drive location having sufficient space. Maybe copy 1 or 2 files and check if they're intact before proceeding with recovering the entire thing.

TL,DR RAID metadata gets messed up all the time because of intel's carelessly written firmware. If ur RAID volume is broken then disable VMD, use Reclaime to analyze the individual drives and dump reconstructed RAID into a VHDX file. Mount VHDX in windows and use TestDisk to recover data.

There will be several prompts along the way asking u to initialize the drives or format them. DO NOT DO ANYTHING SUCH THING.

[pranza]

i have also panicked once as i've found the pc powered off and when powered on, one drive seemed gone (using 2x 4T 990Pros in raid0)

without deeper investigation i took them out of the pc to an external usb enclosure and hooked to another computer - both got recognised, showed 100% life and overall fine!

so i've put them back, also tried swapping places too. 

eventually noticed what you also noticed - that one branch (the one that's not direct to cpu) has unmapped itself from VMD - as soon as i re-mapped it, the next boot my raid0 was visible again!

so the good part is that you don't have to do much more than re-map the second nvme every time that happens

but the bad part is that it happens again and again, mostly when doing backups. (is high i/o to blame?)

so before you mess with the data recovery make sure you've re-hooked the controller back to vmd and of course don't delete the raid too soon - it might be still very well intact!

i'm not sure if it's samsung or intel, but i gotta switch to single nvme.

my rst version in uefi bios is 19.5.0.5676
for the windows 11 driver i tried 19.5,  20 and also 21 - no difference in that regard...
cpu is 13th gen 13900E

[curltron]

Had the same issue after updating my BIOS. Suddenly the first drive (of two) in my RAID0 config was showing outside of the volume.

I wasn't able to just "remap" it but maybe @pranza can elaborate in the exact steps here. Since, I didn't see a way to move the ousted drive back into the RAID0 volume, I instead (in the BIOS) removed the remaining drive from the volume, and recreated the volume with the exact same parameters. Unfortunately, upon booting, I had an empty drive after trying to assign a drive letter/file system format in Disk Management (maybe I screwed something up here?).

Either way, all was not lost because I had previously followed @Letni_sux_big_time 's advice to backup first, and proceeded to recovery.

So all in all, here are the rehashed steps I took (mostly copied from @Letni_sux_big_time 
1. Ensure you have the same problem: Check that it shows up in the BIOS, but one or all drives are missing from the RAID array.
2. Either follow @Letni_sux_big_time 's advice and totally disable VMD (I couldn't boot tho) or use Intel RST (windows app) to remap the drives so you can at least see them in disk management (DON'T try to recreate RAID yet or anything that could cause a format).
3. Download Reclaime RAID Recovery (not just Reclaime Recovery, which I did at first. They are different!)
4. Select the previously RAIDed drives from from the list and select the previous RAID type (RAID0, RAID1, ..., etc)
5. Reclaime RAID Recovery will detect and reverse engineer the RAID parameters
6. (I took the free route, cause Reclaime is stupid expensive) In Reclaime (post analysis) dump it as a windows vhdx virtual hard drive format on another drive that has the space.
7. Mount the VHDX file (NOTE: you won't be able to browse it yet!): Go to This PC->Manage->Storage management ->Actions->Mount/Attach VHD. Again, do not expect to see anything in "This PC" or "My Computer" yet.
8. Make a copy since you can ruin this in the next steps
9. Download and run TeskDisk
10. Select the mounted image (In my case it was named "Msft Virtual Disk")
11. Select "EFI GPT" as partition table type
12. Select "Analyze", then "Quick Search" to find your partition (I had one, but you may need to do a Deeper Search if you have others it doesn't find)
13. Select the found partition and press ENTER
14. Use the left/right keys to select "Write" which will MODIFY your .vhdx image to be readable in windows (for faster copy)
15. Upon completion, unmount the virtual drive in "Disk Management"
16. Double click the .vhdx file you just modified and it should mount in "My PC" / "My Computer"
17. Delete the old/Recreate the RAID (I did it in BIOS) (VMD will be on again, and you may benefit from initiating remapping in the Intel RST app first to ensure your OS drive will boot normally)
18. Upon reboot, go into "Disk Management" and assign drive letter / reformat your Raid drive as it was (it will be wiped clean)
19. Ensure your .vhdx backup is mounted (mount it by double clicking the file if it's not)
20. Begin the copy

Now, you can directly recover straight from TestDisk as well (after having it list files), and this maintains all the original metadata, but I found it way too slow. Modifying the partition in the .vhdx file enables the much faster Windows file copy method I used here (but you need enough HD space for two copies of the .vhdx file as backup).

[pranza]

hi!

well, you go to the bios (efi) and see that raid has degraded - oops, one member missing
then you notice that the whole controller where drive is attached is dangling by itself, with single drive, out of VMD pool (how to say - in 'native' mode);
then you go to VMD configuration and add it back (something like - map this root controler under VMD).
then system must be power-toggled, and just to be sure go straight to uefi and verify that both controllers are under VMD and also your raid status should be showing "normal" with both members present!

in my case i had my system volume on raid0 so there was no chance to boot dysfunctional raid and mess it up from within Windows, hehe. all i did was mangling in bios (efi) and i did not delete raid or re-initialise members. 

by mangling i mean making sure both needed controllers are mapped under vmd - and then raid is back to 'normal'.

your case might be different as it happened after bios update

a recent windows update to 24h2 messed up my adaptec raid's partition table though!
was able to make it work again with help of paragon partition manager and testdisk (copying such amounts of data was not very much an option - too much, too long, where to?..)

you might see from the pictures how awkward it was - first boot after failure it wouldn't even show the abandoned controller anywhere!
only second reset would bring it up in the list and then i could 'enable' it and the raid would be ok again.