Hi,

We received a WD Sentinel DX4000 NAS storage with 4 x 2 TB that the customer forgot to switch off before a planned power outage, and which now refuses to start.

A priori, the NAS enclosure itself is working.
The four drives initialize and a small blue diode appears in front of each drive when it is ready.

The small display panel of the NAS displays: and then We checked the S.M.A.R.T. values of each drive. They were ok, excepted for drive #3, which had a few bad sectors.
We cloned it to a healthy drive of the same model (WD RE4-GP WD2002FYPS).
The cloning revealed that the amount of bad sectors was about 67 Kb.
We swapped drive #3 with its clone, but this did not suffice to restore the system.
We also tried starting the system without the drive #3.

According to the Administrator's Guide, http://www.wdc.com/wdproducts/library/UM/ENG/4779-705064.pdf, the system supports RAID 1 and RAID 5. The customer doesn't remember which kind of RAID was used, nor the total storage capacity.

The partitioning sheme of drives #1 as well as drive #2 is as follows: When drives #4 is attached to another computer, whilst drive #3 is, it is not "In line" but "Off line" in the disk manager.
But when attached alone, it is becomes "In line". So, there is some conflict between the two drives, as they were same.

Drive #3 as well as drive #4 have a unique RAW partition which covers all the drive.

The computer technician who usually cares for this company believes that the system was in RAID 1.
The identical partititioning shemes of drives #1 and #2 could suggest this.
However, the manual suggests that the maximal RAID 1 storage capacity for this system is 2 TB, and this is also in contractiction which the size of the third partition (5528.78 GB), which suggests that at least three drives are used.

So, my assumption is that there might be some RAID 1 by default for the system partition, on which Windows Storage Server 2008 R2 Essentials should be installe, as there is such license sticker at the back pane of the NAS, and then from within Windows, the description of some logical drive, spanning over the remaining space of all drives, or alternatively RAID 5.

However, what I find strange if the system partition was in RAID 1 is that no file system is detected.
As this NAS is running Windows, I expected the system partition being NTFS or so.

There is some reset button at the backpane of the NAS, but I don't want to play with this, unless all drives were cloned first.

The manual suggests that the user can unlock the storage using a password.
I hope that there is no encryption, BitLocker or else.

The goals are :

1) Recovering the data
2) Restoring/reinstalling the system, if possible with the data, otherwise without the data, but after their recovery.

Does anyone here had some experience with the WD Sentinel series ?

Should I assume a RAID 5 system and try reassembling the third partition of drives #1 and #2 with the unique partition of drives #3 and #4 ?

Thank you.

I usually forget such technical details rather quickly, but if I'm correct, WD Sentinels can work in RAID1 mode only if it's a two-drive model or if there are only two drives inserted.
Moreover, your description (if everything's correct) suggests that you're dealing with RAID5.

If nothing else has failed or has been done by the client, cloning the drives and recovering them as RAID 5 should work smoothly.

Start with UFS Explorer, I believe RAID Recovery edition (as well as Professional) will reassemble it automatically:
http://www.ufsexplorer.com/download_stdrr.php

Should anything go not as planned, I'll be happy to help recovering the data from this NAS remotely.

_________________
• Remote RAID, NAS, SAN, VMware, DVR (CCTV), flash and tape recovery. Data recovery support.

Thank you Dmitri,


The manual suggests that the DX4000 can also work in RAID 1, but with a maximal storage capacity of 2 TB.
I don't know if in such case drives #3 and #4 are ignored or if the mirroring is done on as many drives are present in the NAS.

Yes, I also assume the system is using RAID 5.

RAID Reconstructor from Runtime is an alternative to UFS Explorer to reconstruct RAID 0 or RAID 5 : http://www.runtime.org/raid.htm

Connect All Disks and open UFS and send us here screenshot

Prefer if you mark those drives by anything using MS-Paint or something

i had similar Raid5 case last month and did it

_________________
Kuwait Data Recovery - UNIX GTC
The only reason for time is so that everything doesn't happen at once. By: Albert Einstein

Thanks einstein 9,

So, here are the screenshots of the disk manager and UFS RAID Recovery.
The docks in which the drives are were powered on respecting the order of the drives had in the NAS (1, 2, 3, 4).

I also did some analysis of the drives with an hexadecimal editor.
The MBR (signature 33 C0 8E D0 BC) was located at the following offsets:
[list=]Drive 1 : 7B55A90
Drive 2 : 7ADDA90
Drive 3 : 0
Drive 4 : 0[/list]
(Drives 1 and 2 had "00" hexadecimal values at their beginning.)

I also searched for NTFS boot sector patterns (signature: EB 52 90 4E 54 46 53 20 20 20 20).
They are given by the screenshot of the spreadsheet.

And this is where I stop for today.

You might want to look more at the hex rather than the manuals. You should be able to tell if the drives are mirrored by comparing hex values at different parts of each drive (especially drives 3 and 4). If the drives do not match then its most likely RAID 5. If two sets of drives match, then its a RAID 1.

_________________
HDD, SSD, Flash and RAID Data Recovery
Founder of The Data Recovery Professionals Group

@SOSdonnees

This is Raid5

_________________
Kuwait Data Recovery - UNIX GTC
The only reason for time is so that everything doesn't happen at once. By: Albert Einstein

Indeed, I didn't see the last post. A 5.5TB partition. I also meant 1+0 rather than RAID 1

_________________
HDD, SSD, Flash and RAID Data Recovery
Founder of The Data Recovery Professionals Group

Raid5, Backward dynamic, 64Kb stripe size.
In sector 3907023920 of each drive should be present interesting info about this raid5 layout (Order, stripe size, offsets and so on)

_________________
Angel Data Recovery

hehe As a Conclusion here:

From your Uploaded Image, Disks Order as the following:

Disks 1&2 (One of them is 1st. the other is the LAST) in array

Disks 3&4 (i guess should be simple, 2&3) in array

The rest is you to collect all info.

Good Luck

_________________
Kuwait Data Recovery - UNIX GTC
The only reason for time is so that everything doesn't happen at once. By: Albert Einstein

I totally agree. However, the documentation helps looking in the right direction. The manual only mentions RAID 1 and RAID 5. So, useless to search for RAID 10 (1+0). If the system supported it, it would be mentioned as it would be a sales argument.
Absolutely. The last screenshot in my second post shows they are not mirrored.
Totally agree. Yes it's RAID 5.
With 4 x 2TB drives, RAID 10 (1+0) would allow a maximum storage of 4 TB, not 5.5 TB.

I ran a XOR test with graphical output, showing that the RAID 5 is OK for most sectors.
The first 3 sectors failed, showing where the problem could come from.
I also observed some small stripes at some locations, where the XOR test was ambiguous, which proably means that the XOR test succeeds for some bits in a sector and fails for some other bits.

----------
@DR-Kiev:
Outside the fact that the 64 Kb stripe size is widespread, how do you know that this RAID 5 uses 64 Kb stripes and that it is "Backward dynamic"?
Did you ever work on this NAS model ?
In sector 3907023920 of each drive are only zeros.
----------

Data recovered

Drive order : 2, 4, 3, 1
Block size: 128 sectors (64 Kb)
Backward

I worked many times with them, this is default settings for this NAS . Users very rare change it.

_________________
Angel Data Recovery

Hi again,

Now that the data were recovered and saved in several locations, how could I put the WD Sentinel DX4000 back to work ?
I want to restore it to factory settings or --better-- to repair the RAID-5 of the data partition containing the files or reinstall the Windows Server operating system ?

The RAID-5 data recovery that I did was read-only, so that the RAID-5 must still be repaired if I want to use the NAS again.

A XOR test showed that most of the RAID was healthy, some locations of the RAID-5 were "ambiguous" and a few ones were defective. The drives themselves are now healthy as the defective drive was cloned and replaced.

In the documentation of the system, I only could find the how to hot swap a defective drive and let the system automatically restore the RAID 5. But as the drive that was cloned contains data that can help to reconstruct other sectors of the RAID, I'm not sure that this would be the best things to do.
All depends on how the rebuild of the RAID-5 was implemented.

When the RAID-5 system tries to initialize, the displayed error is :

Hi there i have exactly the same problem, the sentinel dx4000 with 4 drives of 2Tb give me the error:

STARTUP
FAILED: 0xD9

I need to recover the data, I already have the 4 drives connected to USB ports on a laptop, but i need to recover the data, what software did you use SOSdonnees , can you please help me out here.

Thanks

Fixing the Boot issue is a headache, you will waste alot of time repairing this Storage (took me few days to get it back online)
i suggest to recover your files to another NAS (recommend Synology)

good luck

_________________
Kuwait Data Recovery - UNIX GTC
The only reason for time is so that everything doesn't happen at once. By: Albert Einstein

Hi there thanks for replying, i really need to take the data out, its about 5 TB of projects that this NAS has, has i told you i have the 4 drives connected via USB, i now the positions that the drives were on the NAS, I have the UFS Explorer to build the RAID, but i really need to know if i´m doing this right to dont brake the DATA, Its critical
Can please someone help me out.

@pgeraldes

The solution is very clear here (If you a real DR Company),
but IF not am sure there are MANY DR companies ready to Assist in your country (Paid of course).

_________________
Kuwait Data Recovery - UNIX GTC
The only reason for time is so that everything doesn't happen at once. By: Albert Einstein

I´m not a recovery Enterprise, I´m the IT Responsable of my enterprise, my main area is networking, I thought this forum was not exclusively for HD Gurus, but thanks for your help.

@einstein9

All Data recovered thanks for all the help