Hello all,
I have a Western Digital WD15EADS-00P8B0 1.5TB (32MB CACHE) (with NTFS uncompressed) that runs at excruciatingly slow speeds as less and less space is left on the drive. To use it (kinda) efficiently I need to leave more than 100GB free, and even then, copying speeds are quite low. As I fill it up and reach the 30GB free space, any file copy takes hours and even errors out. These green drives are dreadful and what I'd like to know is whether there is some kind of format procedure or factory settings tweaking that I could do to make it a usable drive. Any tip would be appreciated.

What you are describing is abnormal and suggests the drive is probably faulty. More details are needed in order to try to better understand the problem, including:

What is the history of the drive - any mechanical shocks or other unusual events?

Collect the full raw SMART data, for review (do NOT perform any SMART or other disk tests).

What are the exact errors that you mention during file copying?

Do you need the data from this disk or are you just looking to make the disk reliable and the data is already backed-up elsewhere?

This drive is used as a backup drive, just storing files that aren't accessed frequently anymore, and never leaves the house. It's almost always turned off in its USB enclosure and I use it only when I want to copy files there, not as a live, always connected to PC drive. No falls or shocks up to now. When I go home I'll look at the SMART data and post them here.

File copying starts normally, and then after a short while data transfer speed drops and drops and drops and if you leave it copying it even disconnects and you get a device not found from Windows. From what I read around it seems many WD green drives like mine had a bug in their firmware and were practically useless (imagine trying to run an OS on top of that mess) but I haven't found any solution to that no matter how much I googled.

I need the data from this disk and will try to empty it before executing any changes (that will take a while...)

connect it via USB and check the events viewer , sure you will see bad sectors, if that is the case, you will need to seek a pro to help you
good luck

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

@masbass,

Thanks for the info so far. Here are my comments (all IMHO) based on what you've said. Other people may have different views.


Without wanting to be pedantic, since you say that you need files which only exist on the drive, that isn't what most people would call a backup (i.e. it isn't another copy of data which exists elsewhere also). The point is that this news changes the whole risk/reward ratio, if the data on this drive is important to you. Any DIY recovery attempts may cause further / faster deterioration of what is likely to be an already-failing drive. If you are lucky, you might recover most or all of what you need; if you're unlucky or make mistakes, then you won't and your attempts may make any later professional recover harder and/or more expensive or, in the worst case, impossible.


OK, so no hints there what may have triggered the problem.


Thanks - remember, the full, raw SMART values are needed, in addition to the usual "value, worst, threshold" SMART values for each attribute. As einstein9 mentioned, some unreadable sectors are very likely - I'm interested to see that info (SMART attribute #5) and other attribute values, to see how severe the drive thinks the current situation is.


OK, that still fits with a possible drive fault.


Your problem description is so severe that I don't believe you have a f/w fault. However if that is the only path you want to persue, just say so and I'll leave you to it and wish you good luck.


As I mentioned above, DIY recovery attempts have risks, which you need to consider and accept before you continue.

You have 2 basic options to try to get the data off the drive - either try to copy the files (if you only need a very small number of files, you might be lucky doing this although you've explained that your drive won't successfully copy for long), or cloning the whole drive to another empty, working, 1.5TB or larger drive, using software which can cope with (and has some error recovery control regarding) unreadable sectors. There are a few choices for cloning software depending on your budget (some are commercial, some are free), your expectations regarding features (GUI or CLI, supported or not, degree of control etc.), your skills with other OS (some are Linux/Unix, some are Windows, some are DOS etc.). In my (non-DR) work, if I need to do this, I use GNU ddrescue on *nix, but that is unlikely to be suitable for you unless you have some *nix experience, and even then, it is not a "one click" utility. Other previous recommendations from other members include DMDE, Media Tools Pro, HDClone etc. Do not try to use Ghost, Acronis, XXLClone etc. as they are not designed specifically to cope with failing disks. You must get the direction of the clone (source vs. target) correct, otherwise you irreversibly erase your data!

Unless your USB enclosure (which I guess was one you bought and fitted the drive into) does encryption, then in general it is usually more successful to clone when the failing drive is removed from the USB enclosure & attached via (S)ATA instead of USB. The (empty, working) 1.5TB or larger target drive of the clone can be USB-attached, since we don't expect it'll need error recovery to occur during the clone. If your enclosure does do encryption then this needs to be investigated further to decide on your options.

Overall, unless you need only a few files, then cloning the whole drive (or at least the partitions containing data, is that is much less than the whole drive) with suitable software is generally recommended - but you're still taking a risk whatever you do. (I believe that DMDE has some features in this area to minimise what needs to be cloned, assuming that the MFT is readable, but that isn't software I've investigated myself, as I can't use it at work.)

If the data has more than minimal value to you, you may want to consider using the services of a DR company, if they have dedicated cloning devices (many companies do), as those devices will typically do a better (more successful & less risky) job of cloning a failing disk. Your data, your risk, your choice!

Thanks for your replies, it seems the hard drive is performing better when there's more empty space left. So, it would take some hours to unload the first 50-100GB, clear some space by deleting the salvaged files and then the next copy will be faster etc. I just need to do the first files in small 5GB chunks (no problem there as most are already large files) and they will hopefully be transferred at a reasonable rate. I can access the drive normally and open / view all files I randomly try, it's just excruciatingly slow and from what I read on the internet ( http://bit.ly/HO03Z0 or http://bit.ly/HO0liI or http://bit.ly/HYIhAY ) it seems this is a known issue among many WD Green drives (no solution found though). Here is the SMART data

Oh yes, it's known all right - just not shown to be a f/w problem, which was what you said earlier on. We're agreed there are similar problems (with varying degress of severity, and varying results to "workaround" attempts) being reported by many users


Thanks for that. Disappointingly, I see nothing abnormal, but as always - lack of problem seen in the SMART data does not prove that no problem exists.

There are other diagnostic tests which could be run, but getting that data off the disk should be the priority IMHO, and I've already given suggestions about that (which didn't include doing file copies for large amounts of data, as you've decided to do). Good luck

hmmm... problem with cloning is that I have no 1.5 TB disk available. My brother can lend me a 1TB he has and if I delete around 600GB duplicate files (files I have already stored elsewhere as well) I might be able to free enough space and make it work. The files inside this HD are not unique / personal / work related, they're mostly stuff like documentaries, tv series etc that I can find again, so it's not a high importance file operation. I'll look on those windows/dos utilities you mentioned above and ask around for a 1.5TB disk (the way it works -if I remember well- in order to clone a disk one needs to have another disk with same size, even if the files to be salvaged take a lot less space ). Prices for new HDDs are way too expensive still so as the files are of minimal importance I'll try the cheap way. Will come back when disk is empty, thanks for your tips!

Ok, copied all data off the disk successfully, same pattern: at first, as the disk had about 30GB free it was copying at horrendously slow speeds, but as I was deleting the files that were copied away it went on it real fast. It is empty now, what checks do you want me to perform on it? I have downloaded Victoria and WD Data Lifeguard Diagnostics

Scan it with MHDD.

_________________
data recovery Poland

To me, this sounds like a logical or file layout problem as opposed to any impending doom with the mechanicals or electronics.

_________________
On a clear disk you can seek forever.

@Keatah,

Based on my experience, that hypothesis does not fit with this statement from the OP, which was in answer to one of my earlier questions:


but I'm interested to hear your explanation for how a logical (filesystem) problem causes that behaviour.

IMHO the results of the MHDD scan, suggested above, will be interesting although I'm not prejudging how conclusive they will be.

Apologies for not reading the details in the thread completely. I spoke before I thought it out completely too. I had the immediate idea the drive is thrashing all over the place, like being badly fragmented. Imagining the $MFT was in 200 pieces and each file triple that.. And so on and so forth.

But that's not likely being a "data-only" non-boot drive, isn't it..?

_________________
On a clear disk you can seek forever.

Understood - however as you now see, we have the OP's answer to my question providing evidence against it being that type of problem. Hopefully the results of benchmarking the drive itself (e.g. with MHDD Scan or similar) will give some food for thought.

Thanks for your answers, I'm trying to run MHDD on a small HDD in order to check the problematic drive as I need to somehow save the LOG files somewhere. The PC has no floppy and running MHDD from cd is read only. Is there a way to install MHDD on an HDD? I only find cd / floppy images and a copy of MHDD I installed on a FreeDOS partition doesn't see the disk at all (only two other SATAs plugged in next to it). Will try and switch cables (although the hdd I m running MHDD from is IDE and rightfully blocked from scanning).

Edit: something I noticed on the surface test MHDD was performing, there were distinct strong diagonal patterns all over the drive as the drive map was scrolling down. I did some tests on my other drives (and one was found to be in really bad shape and will be replaced) and I hadn't seen those stripes as evident as on the one I'm discussing above. Could this be a clue of some sort? Strong fragmentation maybe? Wouldn't that be fixed by formatting the drive and recopying all there or are file fragments copied along from HD to HD?

Since the suggestion to use MHDD was given by laptokowiec and not me, I don't want to "step on his toes" with whatever he will reply, now that you are following his suggestion so I'll just make a couple of brief comments:


It is normal for those patterns to vary between different drive families.


Fragmentation is a filesystem concept. MHDD accesses the raw drive, bypassing (i.e. ignoring) any filesystem, and so filesystem fragmentation (if it exists) would be irrelevant.


I don't understand your exact suggestion, but "file fragments" (if any) have no effect on MHDD results, as I explained before.

I don't know exactly how laptokowiec will suggest you to use MHDD & to report its results (there are at least a couple of different techniques), but I hope my comments avoid you spending further time guessing about "file fragmentation" causing the diagonal patterns.

I see, thanks. I managed to load MHDD on a bootable USB stick with FreeDOS and here is the LOG saved there. I did this scan using the default MHDD settings, is there some other set of options you want me to run MHDD with? I saw in the MHDD manual that the procedure is: scan -> erase -> scan again. Any tips are welcome.

so, I erased the disk and reran scan and here is the log:

Each time you do something, bad blocks increase.
The solution do exist but unfortunately you can't do on your own.
Replace drive, RMA it if possible or live with it. Or refer to someone who can professionally refurbish it.
Final note : yoiu bought a cheap drive and what you pay is what you get.

Ah, it came inside a USB enclosure and I had no idea it would be so bad. Wouldn't it be better if I run MHDD again and delete those delaying blocks marking them as bad? I guess some more would appear soon though... At least no system is running from this HD and it's there just to store away stuff with minimal use.