jeffdgr8

I have a RAID 5 setup with 5 drives using the Intel ICH10R chipset. When I flashed my BIOS recently, it reset its settings to their defaults. When it restarted, it rebooted into Windows before I could change the BIOS settings back to the proper RAID configuration. After setting the BIOS settings back to RAID and rebooting, the RAID array failed. One of the drives was detected as a non-RAID drive. The partition for the RAID volume was inaccessible and Windows Disk Management detected it as type RAW. How can I repair my RAID array? Thank you for your help.

raptor_pa

It will take a bit of research into the raid settings for your controller, you will need to know the drive order, strip size, and parity direction at the least. Google Raid Reconstructor to recover the data. At this point you will probably need to rebuild a 'virtual raid' with RR, copy the data out to another disk, then recreate a new array. You may have to rebuild in RR with the failed drive as a 'missing drive'.

jeffdgr8

The problem here is that my RAID array is 6TB (4x1.5TB). I don't have another hard drive that can hold the virtual image of the RAID. I don't know exactly what got changed on the hard drives that messed up the RAID, but it seems there should be a way to restore these settings as the data should all be there and in tact (I hope so at least).

raptor_pa

The question becomes how the devices are written to by the controller. some controllers mark the first drive only and store other info in the cards NVRam, others mark all drives and then compare to NVRam. How much data is actually on the array? If you have the params pretty well down you could use r-studio to rebuild the array and then just extract the data you need to another disk. ( about $80.00 US) Experience has taught me that the more you try to recreate the array in place the more damage you do. I just finished a 3 disk Raid 5 recovery for a customer that was <136 GB on 3 SCSI drives, It took a week and cost them $5000 to get back 25.5 GB of actual data. On the controller it showed both first and second drives as failed. Created .dd images of each individual drive and then rebuilt a virtual array using x-ways forensics. Their attempts to repair in place probably added a day and a half or two days to the recovery tho. Try r-studio to at least get your dta off to another drive.

jeffdgr8

There is about 2.5TB of data on the array. I take it using r-studio that I should create a virtual RAID5. The problem is I can only see the RAID volume itself, which still consists of 4 of the 5 drives, as well as the one drive that has been set to non-RAID status. How can I present the drives individually to the program to create the virtual RAID5?

raptor_pa

R-studio will see each individual drive and allow you to build it as virtual, once you hit it right, it should be able to rebuild the missing information from the missing drive. You will need enough space tho to recover the files to, you wouldn't have to recover them all at once tho.

jeffdgr8

R-studio doesn't see each individual drive. Does this mean I have to reset all of the drives to non-RAID status in order for them to appear as separate volumes to add to the virtual RAID?

raptor_pa

I am going to say probably. Generally I advise against working directly on the array itself, preferring to work on images of the disks, but that is a bit tougher in a 6 TB raid to pull off. Proceed slowly, r-disk will probably take quite a bit of time to scan the disks if you need to do a scan to find everything.

jeffdgr8

I went ahead and reset the remaining 4 drives to non-RAID status and I was able to create a virtual RAID5 and add the 5 drives. I set the block size to 64kb (this I know). But I didn't know the blocks order. The only one of the 4 that was able to find any file system was the left asynchronous (continuous). I began the long task of copying the 2.2TB of data from the RAID array, which will eventually have to occupy pretty much all the free space on every other hard drive I own. So far 11 files have failed to copy with the error "start position of file (1652993507328) is outside filesystem size (1603021307904)." The first number varies while the second is constant between errors. I looked at the partition information and for some reason it has the partition size as 1.5TB (the size of one drive), but the NTFS information has the correct volume size at 5.5TB. It seems to me that it does not want to read a file that is outside the initial 1.5TB of the partition for this reason. At least that seems like the likely reason. How do I get it to know that the partition is in fact bigger than 1.5TB?

raptor_pa

Are you sure the drive order is correct? That may cause the issue.

jeffdgr8

I just placed the drives in order 0-4. I tried some other combinations and it seems that only when drive 0 is first does it detect the 5.5TB NTFS volume. If drive 1 is first, it at least detects a 1.5TB partition still. But every other drive being first it doesn't detect anything. Do you have any tips for how to figure out the proper drive order?

raptor_pa

Looks like drive 0 is the first disk. The others, you could piece together and check jpgs. That is the quickest way to tell if you have a raid assembled correctly. In theory when assembled you would not have to scan the disks for files, but the file system should come into line. Write down each combo you try so you don't duplicate them. Also haven't looked at this yet, but I was talking to Scott on Monday and he said he had done this video and was publishing a paper on RAID it may help you. http://www.irongeek.com/i.php?page=v...und&mode=print

jeffdgr8

Thank you for the link. That was very informative! I was able to extract a couple mp3 files, one that was around 500KB and one that was 2.5MB. Both played through completely without any mixing of songs. This was with the drive order as I originally had it, 0-5. So this must be the correct order. I wanted to test some more mp3s and some jpgs but every other one I tried to extract wasn't able to be recovered because of the error I mentioned above of the file being outside the partition range. This is probably due to how the volume was defragmented. All the larger files were at the beginning of the drive and the smaller ones came later, after this boundary. I ran a scan on the volume and after it completed, it recognized the partition as it should be at 5.5TB. I was able to view the file/folder structure as it should be and also recover files that it was giving me errors on before! It turns out I actually have 3.1TB of data, it just couldn't see some of it before the scan. So now I'm beginning the long task of recovering the data to any and all available free space I have! I'll report back later as to the quality of the data that was recovered, whether it has any errors in it. Thank you so much for all your help! Most everyone else who has responded to my problem has either not known anything about the subject or said this was impossible to fix. Thank you, thank you, thank you!!

raptor_pa

Kewl, glad it's getting data! Pat yourself on the back, you just saved about $5000 LOL...

chucknein

I had the same thing happen to me.
I do not have 4.5TB of space to recover images to.

Is there anyway of getting the "kicked off" drives to reattach to the array?
I know the order is good. They are all new drives and still working fine.

-Chuck

raptor_pa

Is it Raid 5?

chucknein

Yes
It's on board RAID 5 by intel storage Matrix.
4 drives total (1.5TB each)
2 got kicked off the RAID array when rebooting after installing a new Video card.

Thanks

Chuck

neilyos

I'm having the same issue as you guys. ICH10R, 4x 1TB drives, and attempting recovery with R-Studio before I resort to emptying the bank account. Any progress on your part so far?

markddd

may be you can try a getway raid recovery, partition recovery supported, you can try their demo version, the file view function allow you to preview the rebuilded raid data. simple to use