Esta página está atualmente no estágio de tradução.
Desculpe pelo inconveniente.

How can one recover files from nested RAID?

Standard RAID levels can be easily combined in one system to obtain additional benefits, like enhanced performance, redundancy or even both. This technology is referred to as nested or hybrid RAID and gets widely deployed by users whose requirements cannot be covered by any of the conventional layouts. Still, even a nested RAID implementation may appear to be helpless in the face of logical issues that result in the impossibility to access mission-critical files or even a total data disaster.

While basic RAID configurations employ either striping, mirroring or parity to create a storage, nested RAID utilizes two of the mentioned methods in a single array. In fact, the nesting technique allows creating RAID which has other RAID sets as components: several portions of physical disks are organized into RAID units of the same type and the resulting logical drives are joined on the higher level into another RAID. The most common nested RAID layouts are RAID 01, RAID 10, RAID 03 (RAID 53), RAID 30, RAID 05, RAID 50, RAID 15, RAID 51 and RAID 60. The lowest level of the system is denoted by the first digit in the name and the second digit represents the highest level.

Recovery of data lost from such a system is possible, however, before that each of the levels in these hierarchy needs to be reconstructed. The procedure itself is performed from the bottom to the top: first, the arrays constituting the lowest level (indicated by the first digit) are to be built and then the obtained RAID volumes are assembled into another RAID, whose type is designated by the second digit.

The number of intact disks that will be adequate for data recovery depends on the applied combination of algorithms. In fact, the fault-tolerance capabilities of such a RAID correspond to the ones of the standard RAID types. The only difference is that the components of the top RAID are themselves represented by disk arrays. For instance, if we take RAID 60, each of the RAID 6 sets it comprises may lack up to two member disks thanks to double parity, but the top RAID 0 configuration has no parity and cannot be built even without a single RAID 6 element. Other considerations that should be taken into account are:

  • If the number of broken disks in RAID is larger than the one tolerated by the chosen configuration, data recovery is possible only after they are fixed. Such a storage should be sent to a professional service provider. If any of the drives has signs of forthcoming failure, it is advisable to make its copy and work the disk image instead of the physical device. Replacing the drive and rebuilding the system may lead to even more acute data loss;

  • Like with any other storage, data overwriting that may be caused by the operation of full format, system activity or the TRIM command on SSDs make the lost files unrecoverable;

  • Encryption of the nested RAID volume in the majority of instances doesn’t obstruct the procedure, provided that the critical encryption metadata wasn’t destroyed and the user has the correct password or decryption key.

Recover data from a nested RAID set, such as RAID 01, RAID 10, RAID 50, RAID 60, etc.

As long as the array remains operable (even if it is degraded), suitable data recovery software will be able to restore the files that went missing due to a logical mishap. SysDev Laboratories offers UFS Explorer and Recovery Explorer as effective instruments that allow handling RAID units of various complexity, including nested or hybrid patterns, like RAID 01, RAID 10, RAID 50, RAID 50E, RAID 60 and many others. If the RAID metadata is not seriously flawed, the program will read out the parameters of the array and reconstruct the initial configuration. Alternatively, the user can assemble the storage manually from its component disks or their images. Either way, the array is built in the virtual mode, without writing anything to the source disks.

To recover files from nested RAID using UFS Explorer RAID Recovery, you will need to perform the following steps:

  1. Pick out a computer that will serve as a host one during data recovery and install UFS Explorer onto it. It is advisable to choose an incompatible operating system to exclude the possibility of implicit overwriting of the RAID configuration. For instance, for Linux MD RAID you should take a PC running Windows.

  2. Prepare a storage to which you can save the recovered data: an internal disk, an external storage medium or a network location.

  3. Connect all the available disks to the host computer. If it is possible, utilize the SATA ports on the motherboard, as they will provide the highest data reading speed. Otherwise, employ USB to SATA adapters or external enclosures.

    Note: Please refer to the HOW TO section for detailed instructions.

  4. Launch UFS Explorer RAID Recovery. If needed, adjust its settings in the respective pane.

    adjust program settings in ufs explorer raid recovery

    Note: To work with disk images instead of the physical disks, open the disk image files using the "Image file or virtual disk" option from the "Open" item in the main menu.

  5. Check the list of connected storages in the left pane of the main screen:

    1. 5.1.1 If the RAID metadata wasn’t seriously affected, the set will be built on the spot and displayed in the list. You can identify it by the special icon, level, number of disks or a logical volume mounted on it (in case it wasn’t deleted).

    2. 5.2.1 If the program didn’t reconstruct the nested RAID configuration, the arrays of the lowest level are to assembled manually from the member disks (disk images) and then combined into the array of the highest level. RAID 10 will be used here as an example, but the procedure is analogous for the rest of nested configurations.

    3. 5.2.2 Press the "Build RAID" button and add each drive or partition constituting the first RAID of the lowest level to the new virtual RAID. For this you may double-click the storage or use the respective option in the storage context menu.

      Note: For more specific instructions as to building standard RAID you may refer to Systems with redundancy and Systems without redundancy.

      adding drives to raid builder in ufs explorer raid recovery software

    4. 5.2.3 If the array is in the degraded mode, insert a placeholder (or placeholders) to indicate the missing disk (or disks).

      adding placeholder instead of missing component in raid builder of ufs explorer raid recovery program

    5. 5.2.4 Establish the order of disks using the arrow-like buttons and set the valid RAID parameters, like level, parity distribution, stripe size, etc. in the configuration sheet. Press "Build this RAID" upon completion.

      build this raid button in raid builder in ufs explorer raid recovery software

    6. 5.2.5 Repeat the same procedure for each RAID unit comprising the lowest level.

      build lowest level raid using raid builder of ufs explorer raid recovery program

    7. 5.2.6 Having finished with the lowest level, open RAID Builder again, create a new array and add the volumes mounted on each of the obtained RAID sets. Assemble the storage just like you did in the previous steps.

      create raid of previously assembled raids in ufs explorer raid recovery program

      Note: If you don’t know the exact RAID layout, you may try guessing it through multiple reconstruction attempts. Select "Edit RAID configuration" in the storage context menu to open RAID Builder for the array, make the adjustments and build the storage with new parameters.

  6. If the assembled RAID doesn’t contain any accessible partitions, you will need to perform partition search. For this select "Manage partitions or regions" in the partition context menu.

    search for partitions with manage partitions or regions option in ufs explorer raid recovery program

    Choose the "Search undetected partitions by metadata" tool. In the opened window tick the file system type you suppose is applied on the storage and press "OK".

    specify partition search rules in ufs explorer raid recovery program

    As soon as the needed partition is found, you can stop the search and press "Finish".

    stop searching for partitions in ufs explorer raid recovery program

  7. Select the mounted partition and examine its file system. Open several images or text files: if they seem intact, the array was reconstructed correctly. To recover the missing files, scan the storage for lost data using the corresponding button or storage context menu option.

  8. Provide the needed scan settings or keep the default ones and press "Start scan". Wait until the process is finished.

  9. Browse the received results: you may sort files by names, dates, typse, employ quick or advanced search or preview images, texts, documents, etc.

  10. Click "Define permanent selection" and choose the items you want to copy.

  11. Press "Save selection" and provide the destination storage for the selected items.

Last update: 09 de agosto de 2022