Data recovery from Drobo NAS/DAS/iSCSI

We are glad to announce support of data recovery services from Drobo storage devices. Our recent researches on "Beyond RAID" technology now allow supporting lost data recovery from most known Drobo models, including NAS devices, DAS configurations, iSCSI configurations. This also includes support of multi-LUN and mixed storages.

 

Cases that can be covered include "bricked" NAS after failed firmware update, different NAS failures during rebuild, partial RAID metadata loss, drive failures etc. Support of "Beyond RAID" redundancy allows recovering data even from incomplete set of drives (if RAID redundancy allows this).

 

Drobo recovery is only available as "remote recovery service" at fixed price 100..250 Euro per RAID component in initial configuration (lower price is applicable for simply "bricked" NAS after firmware or hardware failure).

 

The service can be requested via general contact form. The data recovery service fee is only applicable for successfully completed data recovery and there are no "analysis" or "attempt" fees.

Major update: version 5.16

Starting from version 5.16 the software now preserves file creation and file modification dates from original file system as well as from virtually reconstructed file system after data recovery scan.

These dates also appeared in file system object properties dialog as object attributes.

 

In Professional Recovery software edition there are implemented as standard few additional RAID configurations:

 

RAID 1E - "shifted stripe" configuration

 

 

1. Sample of RAID 1E pattern with "shifted stripe" on 5 disks



 

This configuration has 1 drive of guaranteed redundancy and can be built using 3 disks and more.

 

 

RAID 1E - "running mirror" configuration

 

 

2. Sample of RAID 1E pattern with "running mirror" on 5 disks

 





The configuration has the same properties like RAID 1E "shifted stripe", however has different data allocation strategy. For even number of drives the configuration can be considered as similar to RAID 10 by structure and properties.

 

RAID 6 with compact right-asymmetric parity rotation

 

3. Sample of RAID 6 with compact right-asymmetric rotation on 5 disks

 

This configuration is now supported with both P->Q and Q->P redundancy order and different kinds of indexes assignment for Reed-Solomon code for data recovery with 2 missing drives.

 

Among other useful features there are:

• Support of "capitalization" attribute for 8.3 DOS file names;
• In-place editor for RAID Definition Language (RDL) in UFS Explorer Professional Recovery.

 

This update is considered as recommended for data recovery and forensics specialists

Software update: version 5.14

This software update is major and includes many new features, problem fixes and reworked functions. Following new features are available since version 5.14:

• Advanced filter and parameters specification for file copying via "Copy with parameters..." tool. This function allows precise specification how and what to copy with bulk data copying operations.
• Supported Apple Software RAID. In RAID editions of UFS Explorer this allows to build apple "software RAID" automatically.
• File name and path now shown in "file already exists" dialog. This extension of "file conflict" dialog now allows specifying precise type and location of the conflict.
• Chunked VMware VMDK files now can be opened using "disk descriptor file". This allows to open "chunked" VMDK files as a single virtual disk even in non-RAID software editions.

Total rework applies to modification and unification of "file copying" procedure. Now it works same way in all software editions.

The following problems were found and fixed since previous release:

• Fixed "Dot issue" at the end of folder name under Windows;
• Better support of old big-endian UFS;
• VMDK partial cluster issue is fixed for ESX 5 VMFS;
• Fixed legacy HFS compatibility issue;
• Software now properly handles "dirty" file allocation B-tree on Ext4 file system;
• Fixed bug in restoring connection to offline network attached drives under Windows;
• Fixed crash in RAID builder if disk is disconnected during RAID reconstruction;
• Removed condition for hang of adaptive RAID5 reconstruction procedure.

The specified bugs may affect many users, so this update is recommended to all users of software version 5.

Minor update: software version 5.10.1

New software update introduces mostly reported problem fixes, which include the following:

• Software crash and context menu issue under MacOS X in certain scenarios;
• Damage of "localized" file names in data recovery scan result in certain scenarios;
• Issue that prevented reading of large VMDK from VMFS file system if defined small file system block size.

New feature in this update is implemented support of Intel Matrix Storage RAID technology. With this update, software can automatically recognize LUN volumes on individual drives as well as automatically recognize and build RAID of different levels. Because of design, to make this feature work the drives must be connected via native interface to make drive serial numbers available to the running software.

Minor update: software version 5.7.1

In this update there are fixed several stability issues, including problems of handling "degraded" RAID 5 in certain scenarios. Other implemented updates include fixing of other reported minor usability or stability issues.
If you have any problems related to operating with degraded RAID 5, it's recommended to upgrade your software copy.

Software update: version 5.7 released

New software version 5.7 introduces new features and several problem fixes. The software is recommended for update to all version 5 software users.
The new features in version 5.7 comparing to 5.6 are:

• Support of large Ext4 volumes (over 16TB);
• Support of new generation of ZFS file system with use of "System attribute" instead of legacy "D-node";
• Supoport of VMFS 5 extension that provides extended block addressing in ".sbc.sf".

Among resolved problems are:

• Fixed read operation from RAID 3 with block size below 512 bytes;
• Fixed IntelliRAW scan bug (appeared in version 5.6) when "fragmentation" flag is enabled.

Please note that users of "Professional Recovery" version 5.6 and below for Windows OS will have to re-activate the software using updated registration code.

Software update: version 5.6 released

New software version 5.6 brings more optimization, features and problem fixes. The most major modification is rework of data storage schema of data recovery "engine" that now allows both to save more memory and preserve file modification dates of found (recovered) files. Obviously this is only applicable to files and folders with available file/folder modification dates and not applicable to "synthetic" files found with IntelliRAW.

New features also include feature that allows to define in-memory caching of opened storages.  This  can significantly enhance performance of slow storages or algorithmic RAID with small stripe size (such as 512 bytes for RAID 5 or RAID 0).

There are also solved several problems, including:

• problem with support of VMware "COWD" format used with ESX/ESXi virtual machine snapshots;
• data truncation problem on large VMware VMDK files;
• data truncation problem on extra large Ext3-Ext4 file system during data recovery scan.

The software is recommended for update. Please note however that data recovery scan results produced with older software versions (5.5.1 or below) can not be loaded with software version  5.6.

Software updated: vesion five and half

UFS Explorer is updated to version 5.5 and gets more useful features. The update is considered as "regular" and thus it is free for all users of previous 5.x versions. New software release is available for download from download section of UFS Explorer web site.

What's new in this update

Windows logical volumes
Starting version 5.5 software recognizes and displays Windows logical volumes for simplified and alternative representation of locally mounted drives. It keeps showing classic disk/partition structure, but extends it with new group containing only Windows logical drives. This makes logical drive identification easier and allows to access data on dynamic disks (software RAID) immediately.
If you do not need this feature, you may easily turn it off for faster software startup.

Automated RAID Reconstruction
In this release it was added support of fully automated RAID reconstruction, including support of complex RAID-on-RAID configuration. Software will now detect known RAID configurations and build RAID automatically just after software start. The feature supports both complete and incomplete RAID with redundancy, running them in "degraded" mode.
It keeps monitoring RAID components and if missing part of "degraded" RAID will appear, software automatically adjusts RAID configuration to include missing component.
Want to build RAID degraded? Software will not modify RAID configurations created by the user.

Faster RAID operations
Total rework of RAID subsystem introduces much faster data access in "asynchronous mode". This allows to speed-up access on "striped" RAID (such as RAID 0, RAID 5, RAID 6, RAID 3 etc.) by several times for storage scan and data extraction purposes. This mode allows to speed-up even "degraded" RAID operations and scan RAID 5 or RAID 6 with missing drives much faster than single drive.

Adaptive RAID Recovery
Both classic synchronous and new asynchronous data access modes now support automated adaptive data recovery feature. If one or more RAID components contain bad blocks software may read only good data and adjust on-the-fly data access strategy to reconstruct missing information using redundancy. This function supports major redundancy methods, including mirroring, parity and Reed-Solomon code.

More RAID modes
The software now natively supports more possible RAID configurations (mainly related to RAID 6 and "vertical" data distributions: RAID 3/4, RAID 6/Vertical XOR etc.). New configurations support faster operation and adaptive recovery. Introduced RAID 1 support allows to use its redundancy to override bad blocks or even run it in asynchronous mode at same speeds as fast asynchronous RAID 0 by applying adaptive re-striping.

Full disk encryption support
Professional Recovery software edition now gets decryption module with support of major types of full-disk encryption algorithms and methods (such AES/AES256 with CBC and different Initialization Vector modes, including null, plain, ESSIV etc.). In case encryption key is known or can be recovered, software will be able to decrypt the volume (including RAID volumes) on-the-fly and thus access the data or recover lost or deleted files.
Decryption module is compatible to Linux device mapper encryption module (dm-crypt) and can be used with encrypted storages (including NAS storages) if encryption key is known.
 

Software update: version 5.4 released

The next regular update to UFS Explorer software (software version 5.4) is the usual, free update for users of version 5 software and is recommended for installation. Users of previous software versions (4 and below) may move to version 5.4 of the software according to terms and conditions of UFS Explorer software upgrade policy. The software is available for download from download section of UFS Explorer web site.

What's new in this update

It mostly introduces new features and enchantments to previous software versions. It also includes fixes for several stability and usability issues. The new key features include:

EFS support
Since version 5.4 all UFS Explorer software can recover also lost or deleted files with applied Windows File Encryption (or so-called Encrypted Files System - EFS). The software will not decrypt encrypted data, however it can restore encrypted file at safe target location. If operating system or target location do not support EFS, the software will save encrypted files in format of backup for later encrypted file restoration with free "EFS Restoration Tool" included into UFS Explorer for Windows pack.

Extension to RAID6
There are implemented more Reed-Solomon code based algorithms for data recovery after double storage failure in RAID6. Most RAID6 configuration are now supported, including Hewlett Packard ADG.

Legacy OS support
For outdated systems running Windows NT4, Windows 2000 or Windows XP before service pack it was introduced special software build marked as "Legacy OS". The software will run on any OS platform starting Windows NT4.0 with Service Pack 6 and later. Some minor (later OS specific) functions however will not work in this software build.

More accurate XFS recovery
In version 5.4 the software will recover files names and directory structure more accurately for better final result.

Data recovery from RAID systems

Nowadays, RAID (Redundant Array of Independent Disks) is widely used by home users for their personal computers as a method to speed-up, extend or add redundancy to the main disk storage. RAID-systems are supported today not only by expensive specialized controllers of corporate servers, but also as part of a disk controller on inexpensive motherboards.

Quite stable cost-efficient solutions like Intel ICHxx-R southbridge chips allow to easily deploy a RAID system on a PC. Unfortunately, even the best RAID systems don't feature ever-lasting performance and can fail any time due to many factors, e.g. software or hardware failures. Choosing a reliable data recovery software with tools specially designed for data recovery from complex RAID-systems you substantially increase the chances to get your data back in full scale. Our products can help you recover data from RAID systems, even if RAID failure caused file system damage.

Non-redundant systems

The term RAID defining (RAID level 0, Stripe) or JBOD (Just a Bunch Of Disks, Span) is quite incorrect . These RAID systems have the following practical uses:

JBOD - extended storage consisting of several disks, even of different size. Each component of JBOD follows the previous one to provide monolithic storage with the size equal to the sum of component sizes. JBOD is supported by most hardware and software RAID chips (e.g. Dynamic Disks under Windows can span among different disks or disk partitions).

RAID0 - stripe set on disks of equal size. The data on RAID0 are divided into 'stripes' of equal size and cyclically distributed among all disks. Such 'stripe' size is usually from 512 bytes and up to 256KB. The purpose of data striping is to distribute a long data fragment among all disks that allows to issue data exchange requests to all drives and at the same time substantially speed-up this operation with parallel read or write. RAID0 systems are the fastest and they use the whole disk space.

Perspectives of data recovery from these systems are obvious: even if one disk drive from RAID0 could not be read data from this component could not be recovered. If such failure occurs on JBOD entire fragment of span could not be recovered anymore. For RAID0 this will affect all data on RAID (e.g. if RAID0 is built on 4 disks with stripe size 16KB, after failure of one disk RAID will have 16KB 'hole' followed by 48KB block. In general, this would mean any file with size over 48KB cannot be recovered).

If one or more disks from RAID0 or JBOD failed stop using the disk immediately and take it to a data recovery laboratory. Only physical drive repair could help recover the data!

In case of a system failure not related to disks (e.g. reset of controller settings, controller failure or damage etc.) it's possible to recover data even in case of file system logical damage. The only thing you need is to virtually assemble the original monolithic storage with data recovery software. For this you need to specify the disks included into RAID, drive order and stripe size for RAID0. Data recovery software will read data from components in the same manner RAID controller does and it will be able to access good files on virtually reconstructed RAID.

Mirrors

Typical 'mirror' implementation is called RAID1. Each component of RAID1 contains the same data, thus information could be recovered from any good RAID component. For RAID1 controller can perform parallel reading to speed-up files read access. This kind of storage has the highest level of redundancy and the best perspectives of data recovery.

Data recovery with an efficient data recovery software doesn't require any other actions except direct reading of one RAID component.

Advanced redundant systems

These systems are a compromise between high speed of storage access, storage size and redundancy. Usually they apply the idea of striping from RAID0 but on-disk data is extended with an extra information (parity information) that adds redundancy and facilitates data recovery or even continues RAID operations even after component failure.

Advanced redundant systems include RAID3, RAID4 or RAID7 (stripe set with dedicated parity), RAID5 (stripe set with distributed parity) and RAID6 (stripe set with double distributed parity). 'Single' parity means that data can be recovered from RAID or RAID system can operate after failure of single component; 'double' parity – after failure of up to two components.

RAID3 and the similar systems use classic RAID0, extended with one more disk to store parity. RAID5 and RAID6 distribute parity among all disks to speed-up parity update process for data write operations.

Data can be recovered from advanced redundant systems if RAID-systems has all disks working, as well as if one (for RAID3, RAID4, RAID5, RAID7) or up to two (for RAID6) components could not be read. If more disks failed stop using the disk(s) immediately and take it to a data recovery laboratory. Only physical drive repair could help recover your data!

If data recovery is possible without drive repair assemble your RAID with a data recovery software by specifying drives (including placeholders for a missing drive), drives' order, stripe size and parity distribution algorithm. The data recovery software will read data from components in the same manner as RAID controller does, thus it'll manage access to good files on virtually reconstructed RAID.

Hybrid systems

Vendors often use RAID-on-RAID configurations to improve performance, add redundancy and increase failure resistance. Generally, such systems are combinations of non-redundant RAID systems, 'mirrors' and advanced redundant systems. RAID10 is the most typical system of the group, it employs several 'mirrors' with 'stripe' over them. Here mirrors ensure redundancy and stripe over mirrors adds read/write speed. Data recovery from this system is not a complicated task: take any good component from each mirror and virtually build RAID0 over it.

More advanced hybrid systems include RAID50 (stripe over RAID5), RAID51 (mirror of RAID5) etc. In order to reconstruct RAID 51 you need to build each RAID5 component, and for RAID50 - build RAID0 over a set of RAID5.

Efficient data recovery software will help you reconstruct a complex RAID-system of any level and will increase chances to recover your data.