I made a mistake and I wanted to ask some others for best recovery options.

Boot drive in home server was a single drive RAID 0 connected to a HP Smart Array P410 controller. I was attempting to add a second drive as a mirror for redundancy, but I made the mistake of adding the drive when the original drive was still set as a RAID 0 Array so the controller immediately began striping to the new drive. I removed the 2nd drive from the array immediately (before the progress was at 1%), and surprisingly the software in Windows allowed me to do this with no error, issue, or additional confirmation, and it did so successfully for 2-3 seconds before Windows crashed as the original boot drive was no longer accessible.

What are my best options for recovery? Any chance of getting the original drive to boot again by itself through any bootable utilities? Any way to restore the structure of the original drive?

Any suggestions would be appreciated.


The P410 is a RAID controller, it abstracts the underlying storage away from Windows, Windows just sees a "HP LOGICAL VOLUME SCSI Disk Device".

The P410 will not automatically expand an existing configured array onto a newly inserted disk.

Automatically rebuilding a RAID1/10/5/50/6/60 array on insertion of a replacement disk is a different task to expanding a RAID array.

There is no method in SSA to remove a disk from an array.

SSA (Smart Storage Administrator, which came after ACU Array Configuration Utility) cannot prevent a disk from being physically removed, it does not know that a disk has been removed until the disk is no longer there.

Although only 1% of the RAID expansion of the RAID0 had started, some of the blocks that make up the logical drive configured on the RAID0 array had been moved to the other disk, by removing part of a RAID0 while the computer was on, the RAID set is no longer valid.

If the P410 was fitted with a cache module, and if the cache module had a battery/flash backup, then if the computer was shut down, the disk reinserted and the computer started up, then it is possible (but unlikely) that the P410 could recover.

There is commercial software such as http://www.runtime.org/raid.htm but because the RAID0 "failed" during an expansion, it is possible that it would not be able to recover anything.

Depending on the value of the data, it may be worthwhile contacting specialist data recovery companies.

  • Pretty surely the only way is to go the "reinsert disk and pray". These are the times when you know you need a current backup... – Zac67 Aug 8 '17 at 19:02

In this case, you really messed up by configuring a RAID 0 stripe instead of mirror.

That's really all there is, unless you want to transform the array to a RAID 1+0 with 4 disks or a RAID 5 with 3 disks. It's a terrible lesson, but I don't think you have much recourse.



Don't now, whether you still need an answer here, it's been 20 days ... anyways:

Adding the second disc made the raid controller immediatelly start to reshape your raid. Even at 1%, the process has already started. Now there are two ways to recover:

  1. The raid contoller itself is smart enough to go on in the reshape process to complete it, if you re-add the disc. Then you'll just have to live with the fact, that you now have a two disc raid0. (Get a third disc, copy the data there and recreate the raid as raid1, copy back the data.)

  2. Fix it manually:

Please note, that your raid controller probably stores it's own meta data on the discs, so if you're doing, what I suggest here, you'll have to copy all data to a third disk (make sure, the data is ok there before contining) and then you can copy the data back from that third disc to a newly created raid array, like in step 1. However, before you can do that, you need to understand, what actually happened:

Let's suppose, your disc contain the data (each letter representing a bigger chunk of data, with it's size being controller specific, you can figure it out, if you need to):


Now, you add disc 2:

disc 2: 00000000000000000000000000...

The controller now starts the reshape:

disc 2: B0000000000000000000000000...
disc 2: BD000000000000000000000000...
disc 2: BDF00000000000000000000000...
disc 2: BDFH0000000000000000000000...

and so on.

As you can see, no data is lost. In fact, you can - even without any meta data - figure out, where exactly the reshape process stopped, when you removed the disc.

Now, you need some dd commands to copy chunks A, B, C, and so on to your third disc until you reach the point, where all data is still on disc 1. Copy that rest with one big dd call to disc three, too.

Then check disc 3 to be valid, clean, contain all data and so on. Do a full file system check at minimum.

After that, just create a new raid array with discs 1 and 2 and copy the data back from disc 3.

You're doing this approach with the discs 1 and 2 NOT connected to your raid controller, but instead connected directly to some interface where you can access the data, that's actually stored on the discs them selfes.

You're using Windows?

... probably not for this task ...

  • Great reply, I was proceeding to start with option 1 as per your suggestions, so first I made a raw DD backup of both drives, which was slow, but done. Now I just reconnected the drives to the raid controller, selected the 2 drives and readded them to a Raid 0 array. That's where I am at. I see no activity on the drives like it is doing anything with the old data as you suggested. I have the logical drive from he RAID 0 array consisting of the 2 original disks showing in Windows Disk manager as 3.725gb unallocated on a uninitialized disk. I have not initialized. How would you proceed? – Ajk Tek Sep 3 '17 at 2:24

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.