Hi All,

Hoping that someone can assist, if they are familar with the drive I'm using. I have a Western Digital My Book Mirror Edition 2TB (model no WDH200002U-00),tThe drive is setup in a RAID0 configuration. I changed the mode (not long after getting the drive), from RAID1 to RAID0 using the supplied western digital drive manager software.

The drive appeared to be working fine up until yesterday, when the performance appeared to be very sluggish on the drive. I ran a full scandisk, and this appeared to detect and repair the errors it found. The event log is scattered with the following error messages "The file system structure on the disk is corrupt and unusable. Please run the chkdsk utility on the volume My Book.".. While there are notifications to say the errors were repaired there are more error messages (as shown following this). As it was late in the day, I powered down my system to look at it today, but I noticed after starting up the computer, the drive is showing up in windows device manager as a raw drive 0% space used.

There doesn't appear to be any odd noises coming from the drive unit (I do deal with systems with dead drives on a regular basis, so I'm familar with noises you would hear from a damaged unit).

Can anyone advise if there is a specific RAID/drive controller in the enclosure, or is the RAID level done on a software level? Would there be any chance that I could take the two drives out and connect them directly to the motherboard of the computer, using SATA connections.

Would it be possible to get windows, to pick up the drives as a software RAID0 configuration? Or do you know of any other mechanisms to get the data back? There's only about 600Gb in use, so the data should be on one drive, I'm not bothered about putting the drives back the enclosure, more getting the data on the drive. If you can suggest any software that would be apprecated.

Thanks in advance

Hello,



Congratulation!

With WD drives this is a highway to hell...



You are lucky, one of your drives have bad sectors at this point.
But the state can run to much more worse and the price will be 4x 5x compared to the actual.



Note: all hw raid is an sw raid in one level in the deep....



You will see only the raw drives, and you can't protect it from the windows random and uncontrolled writes...



No.



No, RAID0 is stripe.
This means your data is distributed on the 2 drive in small chunks.



You need special and expensive solutions + experience to recover this case.
I suggest to you, try to find a pro near you before the bad area will kill the head in the drive!
WD is the most hardest case for head replacing!!!
Don't play with your data if have real value!

Best Reagads,
Janos

One drive probably head crashed and destabilized the file system. Send it to a pro...



If you do this you run a high risk of corrupting the data.

_________________
Buy your friends Toshiba\Hitachi and your enemies Seagate.

Sarcasmo, I'm in Dublin, if you PM me, I'll contact you with some further information, hopefully we can help you for a reasonable fee.

_________________
Data Recovery Ireland

thanks for the feedback guys, much appreciated. Yea was a stupid thing to do leaving it on RAID0.. I know

ironically I was trying to tidy up the data I had so I could switch the RAID mode to RAID1 on the drive. I didn't have enough space on the home computer to hold the data, so was in the process of removing duplicate data in order to make space for the switch back to RAID1 .. Doh!

there is data on the drives, but I'm not in a position to fork out a few hundred euros to recover the data. I've spoken to someone, who recently knocked an external 1x1TB drive enclosure onto the ground while it was running, and their best quote having tried companies both north and south of ireland was 700-800 euros, that wasn't factoring in the drive head and clone drive replacement costs.. (ok, in that case the drive motor was rightly wrecked).

I do work with high-end fibre channel SANs, which is the other end of the spectrum to a 2x1TB external drive unit such as this. And I have done some stuff with data forensics in the past, using block level data recovery, but this has been with single drive units (desktop and notebook drives). As I'm not getting the tell-tale clunking sound of the drive head bouncing about the place, I thought it might be a dodgy controller, or a loose connector. (As these units retail for around 200 euros you're not getting the same quality control levels as you would get on a DELL Poweredge, NetApp or EMC raid controller cards.)

As I've never previously dealt with this type of presentation of two drives in a USB connected-enclosure, I'm more interested to know about how the drives are presented, in the enclosure, to the OS. (I guess it's to easier to ask in advance rather than ripping the enclosure apart just to see). If the drives don't have any fancy controller, I could try swapping the drives into another similar drive enclosure, as I have spare similar 2Tb units in work, ordered in the same batch as my dodgy drive. If it's just some form of SATA-USB converter, I would opt to just connecting into my system, and try to present it to the OS.

I'd be interested to know your thoughts.. or am I simply wasting my time, and your time (to reply) ?? I'm more than happy to spend a few hours trying to get the data off the drives, rather than spending weeks digging out stacks of origional media to restore from.

Thanks again for your assistance

This is not a controller problem, i am pretty sure about it.
Additionally based on my experience, WD usb controllers usually not works with any other drive even if the other are similar WD.

So you are only wasting your time, and much more important, wasting your chance!

PM sent.

Janos

The chance has almost gone by chkdsking it, hasn't it?
Anyway, before any attempt in learning or doing you should image the drive and not poke around on a suspect HDD.
Last week, one of our PCs imaged a Seagate drive for 44 hours at 230kB/s, constantly, without any faults or noise. We let it run as the data was not vital and we decided a slow, but faultless image process may probably be safer than tampering with the drive and risking to get nothing, but fast

You are partially right.
But there is still missing some points:
The most important is the proper diagnose!
There is two case, and two good solution.
But if you swap the solutions, both case can turn unrecoverable.
The one is if the head is weak, and the surface is OK, in this case the slow imaging is the right choise. (or a proper tools for 100% safely replace the bad head.)

The other is, if the head is good, but the media have problems. This case need special care, because a force-based imaging program can kill the head, and you know it is WD....
Not the cheapest if have head problem...

Janos

Hi Guys,
Many thanks for the assitance, and the PM's to provide.

I'm pretty much sure that the problem wasn't faulty hard-ware, and possibly a corrupted MFT / partition table. As mentioned as part of my work I do work with higher-end storage systems, but I found that we had some data recovery software in house - Quetek FileScavenger and I also have access to software such as testdisk / photorec, etc. So I gave these a try, firstly taking the drives out of the drive enclosure and connecting them directly into a workstation, booted up using a LiveOS.

The Quetek file scavenger appears to see all the data that was on the drive, but I think I may have some of the RAID0 parameters incorrect. While carrying out a test recovery of the data, some of the files are showing up "length of offset in runs is too large: 5". On my test runs I'm using an offset of 64, but I'm wondering if the westerndigital software has used an alternative offset size.. would you have any reference to the RAID0 parameters for these drives?

thanks again for your assitance.

I highly recommend that you create images of these two drives onto other media before you start playing with them. You can then use the images to try to re-create the stripe. Imaging to other media will also help you to diagnose if you have a bad sector or head fault on one of the drives.

_________________
Data Recovery Ireland

I am wondering what are u basing ur diagnosis on.

My .02 cents would be to image the drives with a bootable linux distro. Then use a program called Raid Reconstructor from Runtime Software (www.runtime.org) to rebuild the array. RR will test the images with different start sector, drive order, block size and direction of rotation. It will then rank each combo and tell you which is the best fit. The program is free to try, but you have to buy it once you have rebuilt the array in the software and want to export the destriped image. There is even a good video on YouTube: http://www.youtube.com/watch?v=DWNq5rAhZ9Y

Shawn

Thanks sknopp for the info. I had looked at this product, along with testdisk and photorec. It seems that testdisk and photorec aren't specifically designed for RAID0 configurations, but I did find a site where some guys did get it to work, but they did have to re-create the RAID array.. which I really didn't want to risk. I'd be willing to try it again, on a drive set that I don't mind playing about with

I did also try RaidReconstructor, prior to posting to the site, it did appear to all that was needed, but I unfortunatly didn't have a spare 1TB drive available (at the time) to image the data to. I've used FileScavenger before, and because I was sure the drives (or drive heads) weren't damaged, I bought a licence for the product, and I have since been able to retreive the data from the disk. It took some tweaking to get the RAID parameters correct - as the data being extracted was all over the place. Western digital it would appear don't want to answer my question about the raid parameters used by the drive. In the end to get the data to extract correctly I had to set filescavenger to think that the RAID0 array was a hardware raid, which pulled the data off perfectly. I can't recommend this product enough for this type of procedure. I have used it in the past when getting deleted data recovered from a drive and it was invaluable at the time.

For some the data extracted the files are showing up as unknownxxxx.doc/.pdf/.jpg/.mp3 but the actual files are perfect. If there happens to be a small handful of files that are corrupted, then no big loss having to review the files and rename them.

For others that may have similar issues in the future, that it's important to get the symptoms of the data loss firstly (what happened in the lead up to the dataloss). If there's clunking or any noises coming from the drive then you're looking at hardware issue, if there's no noise it's possibly data corruption - whether you want to *risk* trying this is down to the individual.

Thanks again to all that assisted with their suggestions