Server storage failures are rarely a single-drive problem. A hardware or software RAID controller, several member drives, and the file system or database sitting on top of them can each fail independently or in combination. DiskAnalyst treats server recovery as a layered problem — first understanding the RAID and controller context, then the individual drives, then the data structures built on top of them — rather than assuming a standard single-drive procedure applies.
Common Causes of Failure
RAID Controller Failure
A failed or faulty controller card can make an otherwise healthy array of drives appear offline or unreadable.
Multiple Drive Failure
When more member drives fail than the RAID level can tolerate, the array goes offline even though some drives may still be partially readable.
Failed Rebuild
Starting a rebuild on an already-degraded array is one of the most common causes of further, more serious data loss.
File System / Volume Corruption
NTFS, ReFS, EXT4, XFS or LVM structures can become corrupted even when the underlying array itself is intact.
Logical or Accidental Deletion
A volume, partition or set of files deleted or reformatted on a server, with the array hardware itself unaffected.
What We Handle
- Windows Server and Linux server storage
- Hardware RAID controllers (commonly encountered environments include Dell PERC, HPE Smart Array and LSI/Broadcom-based controllers) and software RAID
- Servers commonly encountered include Dell PowerEdge, HPE ProLiant, Lenovo and Supermicro platforms
- NTFS, ReFS, EXT4 and XFS volumes, and LVM-managed storage
- Database files and virtual disks stored on server volumes
- Multi-drive arrays with more than one failed or degraded member
Common Signs & Failure Scenarios
Our Recovery Approach
-
1
Preserve the Original Drives
Member drives are not written to and no rebuild is started before the array and drives have been assessed.
-
2
Identify RAID Context
We determine the RAID level, controller type, and original member-drive order before doing anything else.
-
3
Image Unstable Members
Any unstable member drive is imaged individually in a controlled environment rather than read directly and repeatedly.
-
4
Reconstruct the Array
The array is reconstructed virtually from the images or healthy drives, without altering the original media.
-
5
Recover Volumes and Files
The file system, volumes, database files and virtual disks are recovered from the reconstructed array and verified.
Important: What Not To Do
- Do not force a RAID rebuild on a degraded array before it has been assessed.
- Do not reinitialize or reformat the array or any member drive.
- Do not run chkdsk, fsck or similar repair tools on a volume that is not mounting normally.
- Do not swap drive order or replace a failed member without recording the original positions first.
Devices & Configurations We Cover
Frequently Asked Questions
In many cases, yes. The array is reconstructed virtually from the member drives, independent of the failed controller, once the RAID level and drive order have been confirmed.
Not if the array is already degraded. A rebuild on a degraded array is one of the most common causes of further, harder-to-recover data loss — assessment should come first.
Commonly encountered environments include Dell PowerEdge, HPE ProLiant, Lenovo and Supermicro servers with PERC, Smart Array or LSI/Broadcom-based controllers. Bring the array in for assessment and we will confirm what your specific configuration involves.
A RAID failure underneath a virtualized environment is recovered at the storage/array layer first. See our VMware & Virtual Machine Data Recovery page for the datastore and virtual-disk layers above that.
Often, yes, once the underlying volume and file system have been recovered. The database files themselves are then extracted and verified as part of that process.
Ready to Start Your Recovery Case?
Tell us what happened to your device and we'll advise on the right next step.
Submit a Case →