How to recover data from mdadm-managed RAID 0, RAID 5, and RAID 6 arrays created on OpenMediaVault

• Author:
  
  Vladimir Artiukh
• Editor:
  
  Oleg Afonin
• Updated: 16.06.2026 21:14

In this article we will examine the process of data recovery from mdadm RAID 0, 5 and 6 arrays created in the OpenMediaVault 8.2.8 operating system. We will analyze the most common data loss scenarios: accidental file deletion via network protocols SMB, NFS or FTP, data corruption due to failure of one or more drives, server or NAS hardware failure, as well as software failures of the OpenMediaVault operating system caused by updates or critical bugs.

Contents

• Capabilities of OpenMediaVault
• STEP 1. Number of drives required for recovery
• STEP 2. Connecting drives to the computer
• STEP 3. Working with disk images
• STEP 4. Automatic RAID detection
• STEP 5. Manual RAID construction
• Conclusion
• Questions and answers
• Comments

OpenMediaVault is a popular operating system for building home and enterprise NAS servers based on Linux. To provide fault tolerance and improve performance it uses software RAID implemented via the mdadm utility. Most often users create RAID 0, RAID 5 and RAID 6 arrays, which allow combining multiple drives into a single storage pool and protecting data from drive failures.

However, even reliable RAID arrays are not immune to data loss. Failure of one or more drives, controller failure, filesystem corruption, errors during array reconfiguration or accidental deletion of information can lead to loss of access to important files.

In this article we will describe how to recover data from software mdadm RAID 0, RAID 5 and RAID 6 arrays created on OpenMediaVault. You will learn the typical causes of data loss, how to correctly connect drives to a computer and use specialized software to restore the array structure and retrieve lost information.

Go to view

Data Recovery from RAID 0, 5, and 6 Created on Infortrend ESDS 1012 RC

Capabilities of OpenMediaVault

OpenMediaVault is an open-source operating system designed specifically for building network-attached storage. It is based on Debian Linux and targets home users and small businesses that need a simple, reliable and flexible network file storage solution. The system is managed via a convenient web interface, so deep Linux knowledge or command-line experience is not required for configuration.

The key advantage of this operating system is powerful and flexible disk space management. OpenMediaVault includes built-in support for software RAID management via the mdadm utility, which allows creating fault-tolerant or high-performance configurations such as RAID 0, 1, 5 and 6. Integrated S.M.A.R.T. monitoring mechanisms provide continuous oversight of the physical health of drives, alerting in advance to potential hardware issues.

The main feature of mdadm-based software arrays is their high portability and complete independence from hardware. Because the RAID architecture — in particular levels 0, 5 or 6 — is implemented entirely by the operating system software, such arrays are not tightly bound to a specific RAID controller or the motherboard of your NAS device.

STEP 1. Number of drives required for recovery

Before starting recovery it is important to understand how many drives from the array must be present and in working condition.

• RAID 0 requires at least two drives and is the only level discussed that provides no redundancy. Data is written in sequential blocks across all array drives, delivering high read and write performance. However, the downside of this architecture is a complete lack of protection: failure of any single drive immediately renders the array unreadable and all data inaccessible.
• RAID 5 is a more protected solution and requires at least three drives. In addition to the data itself, the system writes parity blocks that are evenly distributed among all array drives. In case of a single drive failure, these parity checks are used to mathematically reconstruct the lost information based on data from the remaining healthy devices. Thus the array remains fully operational until the failed drive is replaced.
• RAID 6 is the most resilient level discussed and requires at least four drives. Unlike RAID 5, two independent sets of checksums are computed and stored here, allowing the array to tolerate the simultaneous failure of two drives without losing access to data. This characteristic makes RAID 6 an optimal choice for environments where storage reliability is a priority.

STEP 2. Connecting drives to the computer

Drives can be connected to a computer in several ways depending on the available hardware.

The simplest and most reliable option is to connect via SATA directly to the motherboard. This provides maximum data transfer speed and minimal latency during read operations.

If there are not enough SATA ports on the motherboard, a PCIe-SATA adapter can be used to connect four or more drives at once.

For quick connection without opening the case, external USB docking stations or SATA-USB adapters are convenient. However, this option has lower data transfer speeds, which may be noticeable when working with large volumes of data.

STEP 3. Working with disk images

Special attention should be given to working with disk images. This approach is particularly relevant when drives are physically damaged or there is a risk of further degradation.

In situations where it is technically impossible to connect all drives required for recovery to a computer at the same time, for example due to lack of free ports, you can use the method of creating their images.

You can connect drives to the system one by one, create images from them, and then load all resulting files into the Hetman RAID Recovery application. The utility will work with mounted images exactly like with real physical drives, allowing you to fully reconstruct a virtual array and extract the necessary data from it.

To create an image, launch the application, select the desired disk from the list and click Save disk. In the dialog that appears, specify the destination for the file.

Note that the image size will be equal to the full capacity of the source disk regardless of how much data it contains. Therefore, ensure in advance that the target storage has sufficient free space.

Once the image is saved, go to the main menu and select Mount disk. In the list of types choose Raw disk images and load the file.

It will appear in the device list alongside physically connected drives.

If there are several problematic drives, repeat this procedure for each one.

After all array components are present in the system, either physically or as images, the program will automatically detect the RAID configuration and display its structure ready for analysis and recovery.

STEP 4. Automatic RAID detection

Let us consider an action algorithm for the most common data loss scenarios. We will start with hardware failures — complete shutdown of server hardware, failure of internal components or damage to individual array drives. These situations most frequently cause the operating system to stop seeing the array and result in complete loss of data access.

This category also includes critical failures of OpenMediaVault itself — situations where the system does not boot or operates incorrectly after an update or system error. In such circumstances, the only way to access the drives is to connect them to another computer, bypassing the damaged environment.

A separate group is user errors. The most typical is accidental file deletion using Shift + Delete, which permanently removes data without moving it to the Recycle Bin. Although such loss appears irreversible, in most cases information can be recovered with specialized software because the data physically remains on the disk until overwritten by new files.

To address these problems we will use Hetman RAID Recovery — specialized software designed for automatic and manual reconstruction of damaged arrays and safe extraction of files.

The main advantage of this tool is that it allows successful recovery even in the complete absence of the original NAS server or motherboard on which the array was created.

The utility is cross-platform and fully functional in Linux and macOS environments. In this video we will perform the recovery operation using the Windows 11 operating system.

Proceed to a practical demonstration of recovery after server hardware failure using an example RAID 5 array composed of three drives. For RAID 0 and RAID 6 the procedure is fully identical, so the steps described are universal for all supported configurations.

After physically connecting the removed drives to a Windows 11-based computer, first open the system utility Disk Management. Ensure that all drives are detected by the system at the hardware level.

Critically important: if Windows offers to initialize or format these drives — decline this action. Any initialization will overwrite metadata and cause irreversible damage to the structure of your array.

After automatic reconstruction completes the program will display the array on the main screen indicating the type, total size and file system.

To start searching for data, select any drive from this array and run the Fast scan mode. The program will scan metadata and filesystem directories, detect lost items and evaluate their condition.

Upon completion the screen will display all array contents available for recovery, and the preview function will allow checking the integrity of found files. To complete the process, select the required files and click Recovery in the main menu.

Specify the destination on another healthy drive and confirm the action. The program will preserve the original folder structure and file attributes. After copying completes click Finish — the recovered data will be available for further use.

If the program could not automatically recognize the array or the Fast scan result is insufficient, use Full scan, which performs a deeper scan with the ability to manually specify the filesystems used in the array.

STEP 5. Manual RAID construction

Failures during array expansion, deletion of configuration or accidental initialization of drives when installing a new operating system usually result in critical damage to superblocks. Since service metadata gets erased, automatic detection of the storage structure becomes impossible. In such complex cases you need to use the RAID Constructor module built into Hetman RAID Recovery.

This tool allows manually recreating the exact geometry of the array. You can independently specify key parameters: correct disk order, stripe size, data start offset and parity algorithm type. By constructing the array manually the program ignores damaged service information and virtually assembles the logical volume directly from filesystem remnants, enabling successful data extraction even after partial overwrite.

Consider a practical example of recovering a RAID 6 array of four drives where two drives have failed simultaneously or the array was accidentally formatted. RAID 6 was specifically designed for such situations — dual parity allows recovery even when two drives are lost simultaneously. However, successful recovery directly depends on the accuracy of reproducing the original array configuration, because in such cases the program may not be able to detect the structure automatically.

For reconstruction in this situation use the RAID Constructor in manual mode where you must specify disk order, stripe size and parity parameters.

For RAID 5 the sequence of working with RAID Constructor is similar, but it tolerates the loss of only one disk. In the case of RAID 0 recovery after the loss of any drive using software methods is impossible because this level contains no redundancy.

To operate in manual mode the RAID Constructor requires prior knowledge of the array’s basic parameters. In our example the following values are used:

• stripe layout type – Left Synchronous P+Q,
• stripe size – 512 KB,
• sector size – 512 bytes.

These parameters are specific to each configuration and may differ depending on settings used when the array was created in OpenMediaVault. If exact values are unknown they can be determined by trial directly within the program; RAID Constructor allows iterating options and evaluating results before launching a full scan.

In the constructor window move the required drives from the Available disks panel to the Selected disks list. It is essential to strictly preserve their original hardware order, otherwise the logical array will be assembled incorrectly and file recovery will be impossible.

If a particular disk is physically absent or unreadable, use the Add empty disk function. The created virtual empty disk must be placed exactly where the missing drive was located. This allows the program algorithms to correctly calculate missing data fragments from the remaining working media.

In some array configurations actual filesystem data does not start at the very beginning of the disk but with an offset — a certain number of bytes occupied by a service area. If this information is damaged or missing, the program cannot automatically determine where the filesystem begins and requires manual input of this value.

The parameter is called Offset and indicates the exact start position of data on the disk. If it is entered incorrectly or left empty, the program will either not detect the filesystem at all or will display an incorrect directory structure with missing files.

In our example the Offset value is 135 266 304 bytes — this number must be entered manually for each physical disk during array reconstruction in the RAID Constructor.

If some array parameters are unknown, they can be determined directly by trial in the program. The preview window is the guide: correctly selected values will result in a partition with a detected filesystem and a readable directory structure. This is the main indication that the array logic has been reproduced correctly.

To simplify this process the program provides a Detect automatically function. It automatically iterates possible parameter combinations — disk order, stripe size and offset — and determines the most likely configuration.

This mode is indispensable when all technical information about the array is completely lost. The scanning and verification of millions of combinations usually takes significantly longer compared to simple manual parameter entry.

The total duration of this operation directly depends on disk capacities and your PC performance. However, the wait is fully justified by the highest probability of successful data rescue in the most complex storage destruction scenarios.

After entering all parameters click Add — the array will appear on the program’s main screen and will be ready for scanning and file recovery.

Once the reconstructed array appears in the device list you can proceed to scanning. Depending on the cause of data loss — configuration failure, drive failure or accidental overwrite of array metadata — start with Fast scan. The program will quickly analyze filesystem metadata and display the existing folder and file structure.

Before starting recovery use the Preview function to ensure readability and integrity of the required files.

If the Fast scan result is unsatisfactory or the filesystem is severely damaged, run Full scan. This mode performs an in-depth analysis of the entire array space and can find files even in complex cases.

After scanning completes mark the directories and files you need to restore and click Recovery.

In the dialog specify the save path on a separate drive and confirm the action by clicking Recovery again. After the operation completes the program will report successful data recovery.

Conclusion

Despite the high reliability of mdadm-based software RAID arrays used in OpenMediaVault, they do not provide absolute protection against data loss. Failure of one or more drives, damage to array metadata, errors during RAID rebuild or accidental file deletion can render information inaccessible.

In most cases data from RAID 0, RAID 5 and RAID 6 arrays created on OpenMediaVault can be successfully recovered. The key points are: do not write new data to the drives, correctly connect all drives to a computer and use specialized tools capable of automatically detecting mdadm array parameters and restoring its structure.

Timely and correct actions significantly increase the chances of successful data recovery. Regular backups of important data remain the most effective way to protect against the consequences of any failures or hardware faults.