I have a problem with a WD400JB HDD.

The Sector-Area in the range of 29GB to 35GB cannot be read (Reading Error).
The others sectors go fine and have been imaged so far, except some small areas in the beginning of the drive.

I think the drive has only one platter (or max. 2) and max 2 (or max. 4) used heads.
According to my understanding, this means, that all used heads are still "useful",
otherwise the unreadable area should be bigger.

1. Does anybody have an idea, why the mentioned area cannot be read?
I thought about wearing, but I cannot believe in such a "localized worn zone", or am I wrong?

2. Can anyone recommend a software-only solution to delete G-Lists on old WD Caviar HDDs ?

3. What's about sediv? Am I the only one getting virus warnings from the demo version?
Is this still a official and reputable product?

Best thanks for all seriously meant suggestions.

It's probably right where the heads had a crash against the platter. I've seen plenty of cases where there was a localized damage spot like that.

What are you using to image it? Just a software tool, or an actual hardware imager? You'll get a lot more data in a damaged section like that using a hardware tool that can read using a smaller block size. Software tools are generally limited by the OS's default block read size.

_________________
Data Medics - Hard Drive, SSD, and RAID Data Recovery Service Company

@data-medics

Thank you very much for your reply.

But would a head crash against the platter not mean, that the crashed head was broken/destroyed?
Have you seen a case, where a head survived a crash?

Because assumed that the HDD has only 1 platter and 2 heads, would that not mean, that half of the data (about 20GB) could not be retrieved if 1 head is broken. But because most of the data can still be read, I assume that both heads are still useful.

I'm using an own software-tool for imaging with the capability of single-sector-reading.
The tool is not limited to a default block read size. I could read every sector individually, but that would be very time consuming.

Interestingly the HDD skips the area (and also single sectors read attempts in this area) very fast (too fast for my experience), like the whole area would be banned because of the G-List.

SeDiv is packed with Orien. If you upload your SeDiv package to virustotal.com, you will probably find that the majority of AV software will give it the OK. Those that don't will probably alert you to the presence of Orien.

WD Marvel is a cheap firmware tool that should do what you want. Otherwise you could use NazYura's freeware to retrieve your drive's firmware resources, or you could try HDDSuperTool.

Ddrescue is able to skip bad sections of the drive. Perhaps you could use it to image the bad area.

_________________
A backup a day keeps DR away.

i think your drive has only 1 head.

_________________
Data Recovery Pakistan

Thank you very much for your reply.
I have now tested the demo version of SeDiv and the WD-Marvel version I bought a few weeks ago.
I thought that WD-Marvel doesn't work with old caviar models from WD, but I'll test it.

Ddrescue is certainly a good option if you haven't something better.
But my "own" tool is much more comfortable than ddrescue.
But thanks for the recommendation, appreciate that.

I have also implemented a function for very fast skipping of bad areas,
(it's more like hopping over them )
but the problem on this drive is, that it seems like ALL sectors in the mentioned area are "bad".

So there's nothing to read, and that is from my experience very very strange.
Like someone deactivated the sectors by listing them in the G-List.
I must say, that I have not much experience with something like a G-list
My knowledge about that approaches zero and is only theoretically.
Maybe you could assist me if you like (or someone else?).

But till now I have not been able to find indeed any drive, with both SeDiv and WD-Marvel.
I'll try with another IDE-Controller I have, maybe it works with it.
I'll come back later and let you know.

G-List is a list of defects that are found during the use of the drive. Those sectors are "swaped" by a spare.

It would be like this, LBA 1, 2 and 3 would link to physical location A,B and C. when you add LBA 2 to G-List then LBA 2 will link for example to physical location X, the first available spare. What you state doesn't make sense.

The drive wouldn't "skip" a section because of G-List. It would either be bad and couldn't be read (the area) or you might have translator/zones/firmware related issues that prevent access to an area/tracks but that is not likely at all.

To understand bad sectors and defect lists read here :

http://www.hddoracle.com/viewtopic.php?f=46&t=1402

Regards.

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.

I've seen literally hundreds of cases where heads had crashed against the platters but still worked. In fact I'd estimate that half or more of the bad sector cases we see are a result of that. It could have just been a momentary bump that didn't mangle the head, but made a hairline scratch in that spot just enough to prevent proper reading.

I do find it odd that it's not letting you read any sectors in that area, but I suspect it's just due to the fact that you're software imaging. Could be that the sheer number of bad sectors in that area causes the drive to become unresponsive because you don't have the capability with a software tool to disable the read look ahead buffer.

_________________
Data Medics - Hard Drive, SSD, and RAID Data Recovery Service Company

Thank you very much for your reply.
I have read your very interesting "guide" about defect sectors.
Could it than be the case, that the sectors in the mentioned area are "marked as pending sectors", which will be remapped from the firmware at the next occasion?
Could that be the cause for this strange reading results?

If I have translator/zones/firmware related issues, what would you suggest to try next?
Please read also the other replies from me.

After trying another controller I get it to recognize the HDD, but both WD-Marvel and the demo of SeDiv seem not to work with this old Caviar-Model.
I guess this, because I get the error messages with another Caviar drive which should be ok.
I have now made some screenshots, I did not come to the last days.

SeDiv shows the error messages already after "connecting" to the specified HDD.
WD-Marvel shows the error when trying to read the G-List.

Is there any other software-tool out there (maybe a little bit older) which can handle caviar-models?

Thank you very much for the information. I hope you're right

Could the problem (with the read look ahead buffer) be solved by deactivating it with tools like mhdd or hdat2?
Does anyone know VSC for deactivating this on WD drives?

Clone the drive in reverse.

_________________
A backup a day keeps DR away.

He is right

Maybe wdr (I don't know)

Thank you very much for the recommendation.
Can someone please open my eyes with a pm?

+1 ,
He is Damn Well Right Buddy .You Claim To Use a HiFi Sector Cloning Software ,I Sense in Your Posts That Do Not Have Much Knowledge About HDD's in General ,Maybe Its time To Try Something Better DDrescue With a Graphical Interface

_________________
Regards
Amarbir S Dhillon , Chandigarh Data Recovery Labs [India]
Logical,Semi Physical And Physical Data Recovery
Website-> http://www.chandigarhdatarecovery.com

Nobody knows anything. I try my best to learn more and more about HDD's.
But in this specific field you're right, I have not much knowledge about firmware related issues.
But I'm willing to learn new things. That is not meant to be mean

You may believe me or not, but I have tried ddrescue some time ago, when I had problems imaging another HDD with my "own" software tool.
In my estimation I could not detect any profit in using ddrescue.
On the contrary I gained the knowledge that my "own" tool is much more suitable on reading HDDs with many zones of bad sectors.
This is because my tool reads very fast in multisector mode. When it detects bad sectors it changes the reading mode to single-sector-reading.
Often there are only a few bad sectors in a bad block (from my experience).
After the bad zone is read in single-sector-mode the tool again switches to multi-sector-reading and therefore increasing the read speed again.

Of course my tool is not as good as a hardware imager (and nobody has claimed that),
because it cannot for example deactivate heads when reading.
But in my opinion it is much more powerful than ddrescue.
I use it for over 10 years now and have imaged countless "defect" drives with it, but it never hangs.
Only one single time the PC hanged, but I think this was due to the controller.
For example I tried another very good imaging tool, I think it was called "copyr" or something like that.
But this tool often hanged with my setup.

For the most cases, "my tool" is a really very very good tool for imaging.
The best software-only solution for imaging I have tried till now in my life.

I must add, that I call this tool "my own" tool because I have the source code of it and I can adjust it.
It was a very expensive tool in the days it was sold (in the 90's).
Of course the skills of the tool are not due to me, I do not want to make more of me as I am.

But having the tool and it's source code is very very helpful for me,
because I can build up on a fantastic tool which was developed and tested over many years (I think it was about 10 years).

I think there are better software-solutions out there than ddrescue.
For example Media Tools Professional , of course this tool is too old and too slow for nowadays.

But for someone not having a hardware imager or a brilliant software-only-solution like Media Tools Prof,
ddrescue may surely be the best alternative.

OK, now I am pretty sure that my old caviar drive is not supported by WD-Marvell and SeDiv.

I found another old drive in my treasure which is from caviar series but this time from another family.
This drives is recognized "good" and the g-list can be read.

So I assume that I need another tool for reading and writing G-Lists for my other Caviar-Series Drive.
Does there a demo version of WDR exist?
Where can it be downloaded?

- Your drive is an OLD WDC MCU drive (prior to Marvell MCU) and in your case is cyl 32 drive from Eclipse family :

http://yura.puslapiai.lt/files/wd/wdc/E ... index.html

WDR is a ripp-off WDR MARVEL tool and will only work on MARVEL MCU drives. Same as for WDMarvel.

You would need something like HRT, PC-3000, etc to have access to the firmware of those old CAVIAR drives (prior to Marvel) :



Screen capture is for old HRT showing WD JB support / tool.

At any rate lacking those you can use manual method to gain access to SA like explained here :

http://www.hddoracle.com/viewtopic.php?f=59&t=1490

And you can write MHDD script to read/write SA on your drive, and/or you can use Nazyura WD tools :

http://www.hddoracle.com/viewtopic.php?f=22&t=28

Try - Chk_wd.zip

In any case your drive most likely doesn't have any firmware issue. I don't think the sectors would be marked as pending but a simple S.M.A.R.T. check should show if you have any re-located sectors (G-List) and/or pending sectors. Your drive is too old to suffer from a slow issue like the ROYL (Marvel) Slow issue as the MCU is completly different and the way the drive operates is diferent too, like if it were a completly diferent system.

Your problem is just a bad patch of sectors on the platter due to platter damage. You can try to disable the read cache and use hardware imaging tool to maximize the chances to get the data on those sectors.

http://www.hddoracle.com/viewtopic.php?f=117&t=1410

http://www.hddoracle.com/viewtopic.php?f=30&t=1183

http://www.hddoracle.com/viewtopic.php?f=117&t=1499

Regards and good luck.

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.