So I’m being reminded why I despise RAID 0 completely at the moment. A friend of mine from work (who is now a friend out of work too) wanted to upgrade his PC from an old i7 860 to something a bit newer as along with still being on DDR2 he was noticing some real issues in games with it.
He is fairly IT literate and has previously upgraded the GPU in his PC and transplanted the Dell XPS Studio Mobo, Drives, RAM, etc into a new case before but has never built a PC from scratch and wanted my advice on what new CPU, Mobo and RAM to go with.
So I went with a Ryzen 1600, Asus B350-Plus ATX Mobo and some Hyper X DDR4 RAM 3200 8GB Kit. Did all the upgrade today and all went well-ish but have hit a bit of a snag.
In his previous PC he had the 2 drives his Dell XPS originally came with and a 120 GB Boot SSD he added at a later date. What we’ve realised today though is rather than having 1 2TB drive and the other being a 60 GB drive that was hanging around, he actually had 2 1 TB drives in you guessed it! RAID F***KING 0.
Now I thought not end of world, will slot back in his old PC, boot of a live USB and copy the Data off, however his old board is only seeing one of the drives as part of the RAID now and other as a Non RAID member helpfully. Windows as far as I can tell hasn’t formatted one of the drives or done anything odd with it (its actually complaining it wants me to format them so it can use them).
So my question is, is there any way to get the array back even just cobbled together to get the data off of it? In theory I would think possible but I’ve always avoided RAID 0 like the plague for precisely this kind of scenario.
In future I will think to not trust someone ELSE to know how their drives are configured but its never come up before as normally i know how all my PCs are setup like the back of my hand.
Any help anyone can provide would be much appreciated.
It’s likely that Windows managed the RAID, and has some file that links the two together, although this is just a guess on my part as I generally only do anything RAID-related in Linux. Is the 120GB SSD his boot drive for Windows? you could try setting it back up again before wiping out the Windows installation (as in don’t wipe out Windows until you check) and see if it gets recognized. Honestly at this point, it’s a 50/50 shot at whether you can recover anything from the RAID. That’s part of why RAID 0 is only recommended for performance, and not for redundancy. If one drive is lost from the RAID, you’re pretty much out of luck. You might be able to reassociate them and do some math to recreate the partition tables using the sectors, but it would be a PITA, and still might not work.
Good luck!
P.S. If you have to do the math route, booting into a Linux USB and using dumpe2fs might help you find the superblocks with the info you need: https://linux.die.net/man/8/dumpe2fs
go into disk management in windows and see if both disks are being labeled as “dynamic”. If they arent, or one isnt showing up, see if theres an option to “Import foreign drive” or any options to rebuild the array.
I seriously doubt that it was a harware raid, but check in the bios, see if there are any raid options in there. if so, you may be able to rebuild it from there.
Windows Dynamic Drives have following characteristics (as much as I remember them - I moved pass dynamic volumes to hardware raid, now back to Storage Spaces - a full cycle ):
windows home editions do not have it
windows recognizes its own volumes in Disk Manager, regardless of on which SATA/SAS port they are connected.
if one of the HDD were not connected at some point, windows will recognize the volume and show it as offline even after connecting the missing drive (user manual action, I think, was required in Disk Manager)
So, since you mention that
“its actually complaining it wants me to format them so it can use them”
… this suggest that those were not dynamic volumes.
Whats left is the on-board software RAID (e.g. intel chipset) - I assume there are no hardware RAID. Which is somewhat confirmed by
“however his old board is only seeing one of the drives as part of the RAID now and other as a Non RAID member helpfully”
Mobo bios would not see any Dynamic volume as a RAID, for mobo or other RAIDs windows dynamic volumes are just random HDDs.
I think, SATA RAIDs port order do not matter (I’m not sure about that). However, RAID is always configured on specific controller and motherboards even if use only controllers from Intel chip can have multiple controllers (e.g. one SATA3 and one SATA 2). So maybe the second disk is connected to the other controller?
SATA ports on mobo are usually connected in pairs (if not quads) so pair SATA1-2 should be of the same controller, then 3-4, …
It was being done far as I can tell by the onboard RAID controller in the BIOS since it is enabled and shows the array in question and that one disk is a member, I will check Windows Disk Management as well though.
@1ncanus and @jak_ub are right here, although I would bet money the RAID was handled by the motherboard. Try different SATA ports with the disk that isn’t being recognized. If after trying the different combinations the RAID still isn’t recognized, there are options to get the data back with the correct software.
This is what I have used in the past to recover data from RAIDs where the controller took a dump, even RAID 5. Since you only have 2 disks in the array, it should be fairly easy to reconstruct the array in the software. There are other tools out there, that’s just the one I have used for data recovery with success.
I owned a DELL XPS 420 from 2008 - 2012. It came with RAID 0 configured by default via a controller on the mobo. To access the controller configuration you needed to hit [Ctrl i] just after the main boot screen.
I have a feeling it may have been fussy about which mobo SATA headers each drive was plugged into originally if you removed the disks and put them back. Have you tried swapping them around? Have you also checked each drives SMART status to check that one of them has not died?
So, what created the RAID0 in the first place? Seeing as it was an old Dell XPS with an i7 (if I understand correctly), it was probably (an early version) of Intel RST.
If so, you have to put the old system’s BIOS into RAID mode, and the array will probably be detected. But… I’m not sure how well a bootable Linux distro USB drive will be able to access an Intel RST RAID0 array.
I’d recommend getting another drive (not part of the RAID0 array disks you already have) and using that to install a temporary copy of Windows 10. You can just download the ISO from Microsoft. Then install the Intel RST drivers. Then Attach the RAID drives and see if you are first detected at boot-up, and then in Windows. Then copy your data off to yet another drive. Then once your data is backed up, you can do whatever.
I had an experience with an XPS machine of this era. It took me a minute to realize dell had those drives in RAID 0 because I didnt want to believe that a reputable company would do such a thing, for obvious reasons. Once I figured out what was going on I immediately reinstalled the drives and had no issue getting to the data with a ubuntu live stick. I kinda got lucky really.
If you can get into the raid config (ctrl + i during start up I think?) you could copy the information about the raid config and see if you can set it back up with the same settings. In theory the information should still be there. Before doing that though I would check to make sure you are connecting the drives to the exact same SATA ports that were in use before. It should detect them and work again.
Solid advice from @NetBandit and @Adubs. The only thing I would add is the last thing I would try after all other options is trying to reconfigure the array. If you are lucky, entering the RAID config will have the controller recognize a “foreign config” and ask if you’d like to import it. This basicly means the controller detects the drives as part of an array that it understands, just not one that it created. Generally this is not destructive, but I don’t have much experience working with consumer grade on-board RAID controllers so this might not even be an option. If you really need the data, do a RAW image of both disks before you go messing with the RAID config. Then if something weird happens during the import of the config or whatever else you try in there, you can use the images to do data recovery via software.
I don’t advise doing this. Overwriting the array extents is a sure way to make it unrecoverable.
There may be some MDRAID tools in Linux to help, but I haven’t looked.
Tread lightly.
I’ve successfully used Restorer Ultimate multiple times(* to recover (all) files from a broken RAID array. It emulates most RAID setups you can think of in software.
*) (desktop drives used to drop regularly from an Nvidia RAID on a ca. 2007 mobo… learned my lesson)
The only snag is, you have to recover the files from the original disks to some other volume - the program won’t just magically make the RAID array work outside the program itself. Then you can re-use the RAID disks, make a new array and copy the files back there.
I agree with @NetBandit. RAID reconfig is your absolutely last option as it is probably not going to work and is really a hail mary. The original, unaltered disks are your best chance at reconstructing the array for data recovery. I have used RAW clones for this before, but it just needlessly complicates an already stressful situation. I should have made that point clear the first time.
):