[ZoXoR]

Woohoo!! It worked!!
I had a failed RAID 0 setup, because I switched my drives to different SATA ports, then afterwards I had 1 failed RAID and 1 non-raid disk.
I took a picture of the settings for my RAID, found in the Intel Matrix screen, than deleted the RAID and recreated it.
In Windows, I used TestDisk like you said (a guide can be found here: http://www.cgsecurity.org/wiki/TestDisk_Step_By_Step)
And it totally worked!!! I have everything back, without having to use extra disks!

Thanks a lot!
10 minutes of worked and what a result!
Rep+

 

[marchewd]

I am having a similar error that is listed on this thread. However, I am dealing with a RAID-5 non-bootable array. The unknown drive is showing up as an internal drive within the Intel rapid storage manager. The unknown array drive has the same serial number of the internal drive, except the unknown array drive as a :0 after the serial number. Would these same instuctions still work for my RAID-5 array, or is there another way I can restore this array? I have an extra 2TB drive (same as the four drives in the array). Could I set one up as a spare and would that allow the array to rebuild automatically?

Thanks,
Dan

 

[dilbert]

This worked like a boss. Even on a 3TB (GPT MS file table) formatted with ext2 on Windows Server 2008 R2. Fantastic.

For those with GPT file tables (if you have a filesystem bigger than 2TB, you will), I did a Google search "TestDisk gpt" and found some git commits (dated ~2010) which corrected a couple of bugs for GPT. The TestDisk version 6.11.3 (6 May 2009) would have this bug, so I used the Beta WIP version, 6.12. I encountered no problems using the Beta.

 

[Wiremaster]

Yay! I got my RAID 0 partitions back. 70% of the way through the RAID 1.

 

[Licht]

I want to add additional information. Using ZAR (Zero Assumption Recovery) you can recover any RAID without the controller or settings it was set up with. It does this by, as it was named, assuming nothing. I'm on my second round using it after I gave up on fixing this failed RAID5 (2/3 disks read as offline without reason.) The process is roughly the same as using Testdisk. The Testdisk methode, however, is much faster. ZAR takes a lifetime to process.
Quote:

Originally Posted by Old Hippie;7389203
Thanks, I appreciate the share.

Running a RAID0 array without a back-up is akin to playing Russian Roulette with no empty chambers.

I had a four HDD RAID0 for three years before decommissioning it for my Revodrive. Never once a problem. Gotta love single platter WD drives. (4x160GB)

 

[Old Hippie]

Quote:

I had a four HDD RAID0 for three years before decommissioning it for my Revodrive. Never once a problem. Gotta love single platter WD drives. (4x160GB)

Cool!

But my statement still stands.

 

[Licht]

Quote:

Originally Posted by Old Hippie;13291701
Cool!

But my statement still stands.

Oh no doubt. I was just attesting to the outstanding quality of those little drives. Today they are operating in four separate machines... We'll see which one dies first.

 

[defecta]

yeah well im another one who 'confidently' deleted the RAID array only to lose everything. this guide is not without its perils people. on two different PCs

fortunately we havent lost anything as we have daily backups.

 

[ProjektSUN]

Wow, what a PITA this issue is, I almost died thinking I lost all my data. I have kind of a weird setup and I had a couple extra steps that enabled me to recover all the data. I also didn't have to deal with recovery of a boot up drive as this is just a massive storage array for media. My setup is an MSI Z68MA-ED55 motherboard using Intel's newest Z68 chipset. I have an 80GB Intel SSD boot drive, a 60GB Crucial SSD cache drive (to accelerate the RAID array), and (3) 3TB disks configured in a RAID5 array. The RAID disks were configured as one, huge GPT/NTFS partion of 5.5TB's usable space in Windows 7 64bit. Here's what I did to recover all my data:

1. Go into Intel RAID BIOS using CTRL-I and delete all non-member disks. I also deleted the 60GB SSD cache drive that I was using for the Z68's special caching and removed it from the array. Basically, I removed all the disks, and deleted the array, and removed the SSD from caching... a blank canvas.

2. Recreate the RAID5 array using the exact same name and settings as the one that was lost.

3. I booted Windows 7 up and used the Intel disk manager to add my 60GB SSD to accelerate the array. Again, used the exact same settings that I did before.

4. This is what tripped me up. I couldn't get the TestDisk program to properly find my partitions, it kept thinking it was a MAC EFS partition. It knew something was there, but I couldn't figure out how to make it work right, and the user interface was bugging me. I had bought Paragon Partition Manager 11 Personal Edition back at some point and had it installed already, so I decided to give it a try. I ran the partition recovery tool and it found the lost partition right off the bat, but wanted to do a lengthy scan of all sectors on the disk, which would have taken hours probably, so I just stopped the scan, right after it found my partition (within a min). Paragon's partition recovery found the 5.5TB partition and saw that 2.5TB's were in use. Hooray I thought! But I wasn't quite there... when I tried to recover it, it claimed it had succeeded, but when I went into the Windows Disk Manager, it showed multiple partitions, which wasn't correct... for some reason it had broken up the disk into multiple partitions which were limited to 2TB. I held my breath, deleted all the partitions in Disk Manager so there was only 5.5TB's of free space. I then converted the disk to a GPT disk. I went back into Partition Manager and re-ran the partition recovery and once again recovered the partition. Voila!!! Windows Disk Manager now saw the whole 5.5TB partition. The last thing I had to do was assign it a drive letter and do a quick chkdsk, which didn't really find any problems. In any case, all my data is back, and my system is now running fine.

Thank you to the original authors of this thread, you got me started on the right track and if it weren't for you I'd be crying right now.

 

[Zero4549]

I really wish my search found this a couple months ago.. or that someone would have pointed it out to me in one of my threads... *sigh* oh well. Thank you. Will have to keep this for later.

 

[speedtrap]

I have used this tutorial to recover my raid partition. I replaced my boot disk with a different one so I could use a fresh Windows installation to run TestDisk on, after reading the initialization issues that others had when booting to their normal Windows install.

The recovery was a success, and I can access my important files again. But I did notice a weird anomaly. Windows Disk management shows a 129 MB piece of unallocated space in front of my raid partition. Is that normal? Should I worry? It might be some kind of special reserved space, but it didn't show in the past.

The Intel Raid Storage Technology tool shows that the recovered partition is not yet initialized. Is it safe for me to initialize now? I assume initialization only re-creates the parity data, and does not touch any file data?

 

[Kramy]

Quote:

Originally Posted by speedtrap;14473797
The recovery was a success, and I can access my important files again. But I did notice a weird anomaly. Windows Disk management shows a 129 MB piece of unallocated space in front of my raid partition. Is that normal? Should I worry? It might be some kind of special reserved space, but it didn't show in the past.

Probably the tiny boot file partition that Windows 7 creates. It must not have been recovered properly. Oh well. If you've got your data, mission accomplished!

 

[speedtrap]

The raid array wasn't a boot disk. I have a SSD for that.

 

[Kramy]

Were your drives plugged in when you installed to your SSD?

If not, then... I don't know.

 

[speedtrap]

The SSD contains my regular Windows installation. During the recovery procedure, I disconnected that SSD and temporarily replaced it with a different harddrive on which I did a clean install of Windows 7 x64.

The drives from the raid array were connected to whole time.

129 MB is a bit of an unusual size of the hidden Windows boot partition, because I think the default size for that is 100 MB.

 

[flipp]

Hey guys, found this page while desperately searching for a solution..

Getting the non-member disk error after overclocking, don't know how this happened but I need to fix if possible, most of the data is backed up however not the most recent files, ideally I would like to get it all back, if possible.

Am currently attempting this fix, although i have another layer of difficulty to add to this:

Intel raid was running my four storage disks (WD 2TB caviar greens) in RAID0, however the primary partition is encrypted using TrueCrypt. Yes I know crazy to have 8TB of data in RAID0, but this is a requirement for me as I am an editor with a LOT of content to move back and forward between my storage drives and boot disk (ssd), and this is backed up weekly to a NAS setup.

What are the chances of restoring this volume?

By default TestDisk has jumped to EFI GPT (I assume this is due to the size of the volume) and I am searching now.

I think it may have found it, three partitions have come up, should I try and write them?

Absolutely desperate here pulling hair out, a weeks worth of work is a LOT of hours

(

Thank you in advance anyone with advice or comments,
flipp

 

[Alberts87]

My ssd failed (none raid, system drive).. while doing some testing to get it working again member disk 1 out 2 got none member.

The fix worked excellent. Also I was lost on wich port what drive was connected. So I run step 1 twice while ports changed correctly. Indeed within a second it found my 1TB partition.

flipp... If you have a better idea to get it working again then go for that

... If not apply this fix. Im quite sure it will fix your problem. Even if it are 4 drives it wouldnt change the outcome there all data would still be precent IF none of the disk(s) are death. For 4 drives a standalone raidcard wouldnt be a bad idea?! Performance wise and to rule out this kind of buggy shizzle?

 

[flipp]

flipp... If you have a better idea to get it working again then go for that

... If not apply this fix. Im quite sure it will fix your problem. Even if it are 4 drives it wouldnt change the outcome there all data would still be precent IF none of the disk(s) are death. For 4 drives a standalone raidcard wouldnt be a bad idea?! Performance wise and to rule out this kind of buggy shizzle?[/QUOTE]

Hmm it seems the ones it found in quick scan are much smaller than my
8tb partition. One is 2tb. I am attempting the deep scan option now.

Do you think me having the partition encrypted with Truecrypt is going to make this impossible?

Thanks

 

[Old Hippie]

Quote:

Do you think me having the partition encrypted with Truecrypt is going to make this impossible?

I noticed you have posted on several different forums without much luck.

Personally, I think you're screwed because of the encryption.

Sorry I can't be more help.

 

[Alberts87]

Hmm with truecrypt incrypted. Didnt know that. I ques with this trick the partition-table gets restored... And I ques truecrypt incrypts that (would be odd if not?)... So I ques that your screwed....