trying to recover multiple partitions on RAID 0 volumes

[Ronson]

I have dual mSATA SSDs in RAID 0 in a laptop. The integrated Intel RAID controller rejected one of the SSDs breaking the two RAID 0 volumes that were on the drives. Using Reclaime Free RAID Recovery I imaged the SSDs separately and then wrote their data, unRAIDed, to a single drive. However, I thought maybe instead of cloning the data from the single drive back to a new RAID 0 volume I could use TestDisk to recover the missing partitions. I tried marking the dual mSATA SSDs as independent drives, then recreated the RAID 0 array and two RAID 0 volumes. I booted in to Windows running from a flash drive and ran TeskDisk. The two RAID 0 volumes show as empty, nonpartitioned and unformated, in Windows Disk Management. However, seems like TestDisk shows there are partitions already there and I'm not sure why. See the 4th screenshot. I scanned for partitions and found some and tried writing them to disk but received a write failure message. The testdisk log file is attached.











I know I could have changed those partitions marked as D (I guess change them to P?) so that they'd be in the next screenshot as recoverable but I wasn't sure what I was doing at the time so only tried recovering those first two partitions.

[recuperation]

Interesting. But what is the question?

[Ronson]

For those wanting a question: What do I need to do to recover the partitions?

Perhaps the solution to the write error is figuring out why TestDisk says there is already a partition structure while Disk Management says there isn't. Or am I reading the 4th screenshot incorrectly?

[Ronson]

I'm not sure what happened to my attached log file in my opening post. I don't see it there so here it is.

testdisk.zip

(14.75 KiB) Downloaded 682 times

[recuperation]

I have dual mSATA SSDs in RAID 0 in a laptop. The integrated Intel RAID controller rejected one of the SSDs breaking the two RAID 0 volumes that were on the drives. Using Reclaime Free RAID Recovery I imaged the SSDs separately and then wrote their data, unRAIDed, to a single drive.

Did your recovery software write a log?
There must be a reason for the broken RAID.
Any unrecoverable sectors in the RAID?
Your testdisk-log shows a lot of sector reading failures.

However, I thought maybe instead of cloning the data from the single drive back to a new RAID 0 volume I could use TestDisk to recover the missing partitions.

Sounds OK.

I tried marking the dual mSATA SSDs as independent drives, then recreated the RAID 0 array and two RAID 0 volumes.

When doing this, you wrote stuff on the drives to be recovered. Not a good idea unless you know what is affected by this operation.
You should try to recover from your "single drive".

Your testdisk-log shows lots of read errors.
Check the health status of your mSATA-SSDs in independent drive mode with smartctl -a /dev/sdx where x represents your drive letter. Post both logs here.
After the copy operation towards the single drive any unreadable sector in your mSATA SSDs should have increased some SMART counter.

[Ronson]

I have dual mSATA SSDs in RAID 0 in a laptop. The integrated Intel RAID controller rejected one of the SSDs breaking the two RAID 0 volumes that were on the drives. Using Reclaime Free RAID Recovery I imaged the SSDs separately and then wrote their data, unRAIDed, to a single drive.

Did your recovery software write a log?
There must be a reason for the broken RAID.
Any unrecoverable sectors in the RAID?
Your testdisk-log shows a lot of sector reading failures.

I couldn't find a log produced by the Reclaime Free RAID Recovery software which might contain read errors. I think there is a unRAID progress log which can be saved. I didn't save it but I can redo the unRAID process, try saving whatever it offers to let me save, and upload the file here if you're curious and want me to do that. After the RAID drives were analyzed there was a log containing info about the RAID array that was discovered but that didn't seem relevant to what I posted about here- trying to use TestDisk to recover partitions. It didn't mention any issues with reading data from the drives. Actually, I'm thinking the Reclaime program may have just ignored any read errors.

I thought the RAID was broken for some reason, too. However, after some Googling I found posts from other integrated Intel RAID controller users who said their RAID arrays were broken for no clear reason. Or, perhaps they didn't discover the true reason. I don't know. I checked the SMART data of both drives using a different program than what you suggested. Both drives passed. Before experimenting with the mSATA SSDs I tried to make a backup image of the drives using Acronis True Image 2013. I had some persistent read errors and eventually gave up making backup images.

I tried marking the dual mSATA SSDs as independent drives, then recreated the RAID 0 array and two RAID 0 volumes.

When doing this, you wrote stuff on the drives to be recovered. Not a good idea unless you know what is affected by this operation.
You should try to recover from your "single drive".

Your testdisk-log shows lots of read errors.
Check the health status of your mSATA-SSDs in independent drive mode with smartctl -a /dev/sdx where x represents your drive letter. Post both logs here. After the copy operation towards the single drive any unreadable sector in your mSATA SSDs should have increased some SMART counter.

Yes, I've checked the single drive containing the unRAIDed files. AFAICT, the important files we want are intact. I haven't checked every file but the files I have checked are OK. Yes, it is best practice to not write anything to the drive or drives with issues. Since I already recovered the data to a single drive I decided to experiment a little. I was hoping there is a program that would let me re-add the rejected drive to the RAID array or give me more details about why the drive was rejected or couldn't be re-added. I found no program for that. Perhaps re-adding the drive could have been accomplished with a hex-editor but I don't know how to do use that to fix a RAID array. I was hoping to fix the issues with the drives and have the laptop booting Windows again. TestDisk seemed like it might help after I recreated the RAID volumes.

I switched the laptop to AHCI mode, booted Lubuntu, installed smartmontools to get smartctl, and ran the commands you requested.

smartctl-output.zip

(3.34 KiB) Downloaded 643 times

[Ronson]

I think there is a unRAID progress log which can be saved. I didn't save it but I can redo the unRAID process, try saving whatever it offers to let me save, and upload the file here if you're curious and want me to do that. After the RAID drives were analyzed there was a log containing info about the RAID array that was discovered but that didn't seem relevant to what I posted about here- trying to use TestDisk to recover partitions. It didn't mention any issues with reading data from the drives. Actually, I'm thinking the Reclaime program may have just ignored any read errors.

Actually, after posting that I realized I can't redo the unRAID process with the SSDs containing the original data. I changed that when I marked them as standalone drives and recreated the RAID array. I can redo the unRAID process with the image files I made but if there were any read errors then they already happened and won't be revealed by unRAIDing image files of the drives.

[Ronson]

So... I fixed my partition problem. Using Windows 10 release 1803 installed on a USB flash drive I tried selecting and writing the found partitions a few more times as seen in screenshots 4 - 6 except instead of only trying to write the first two partitions I selected (changed to P) all the labeled partitions (EFI System, Recovery, OS, DATA, Restore) which seemed to be the partitions I want. I still got the write error message. I don't know why. I tried Googling about it. I found some generic suggestions about making sure the disk isn't in use already and checking antivirus/Internet security software but neither seemed to apply to me. While reading through the search results I noticed some people were using the Linux version of TestDisk so I tried that too in Lubuntu. I went through the same steps as I did while using Windows but this time I didn't get the write error message. I restarted the laptop and now Windows boots from the dual mSATA SSD RAID 0 volume like it should. I'm not sure what difference using the Linux version or running Linux made but it worked.

[recuperation]

I am happy to hear to you succeeded. Thanks for leaving a feedback!