Hey everyone

now this might be one of those typical problems to you, addressed many times in the forum, I did try and looked through a dozen pages before posting, to no avail.

My internal HD failed unexpectedly a while ago, it shows up in BIOS every other time I turn my pc on, and is mounted and shown in ubuntu (disk utility reports some bad sector) but it fails after a while of file copying, its also shown as unallocated space in win 7, first as ntfs in cmd, then as RAW after a while
now I have already gone through a lengthy check disk in windows,chkdsk/r to be specific that took about 50 hours and didnt seem to fix anything, it started reporting "unreadable data" around 1000th cluster and went on from there.

now I'm mainly concerned about recovering my data, though i wouldnt mind fixing the HD and having it around for misc. purposes.
so what do you suggest, is there a software that can do a surface check in anything under 24 hours? or anyway that I can recover my good data and leaving bad sectors as is ?
I run both ubuntu 10.4 and win 7 on my machine (which is the only one i got so I'd prefer a software that could be run in os mode, so I can go around doing my bizness while its sweating around bad sectors

As always, consider the risks of you performing DIY recovery attempts (you can make things worse!). Search for previous discussions about DIY risks - there have been many discussions. If the data is valuable, you may want to consider using the services of a DR company with a specialised disk imager. Unfortunately you may have already made recovery more difficult by running chkdsk. That is always a bad idea, when the disk is faulty. Also, stop trying to surface scan the disk - you risk "diagnosing it to death" as another new user reported here recently.

Since the drive is recognised by the BIOS (and Linux), then as usual, if you decide to take the risks of DIY, one typical approach is to perform a raw (sector-by-sector) clone of this 500GB "problem" disk, onto either a different (raw) disk, or into a file in another filesystem on a different disk. You cannot have any filesystem from this "problem" disk mounted while you (try to) perform that cloning, so forget your plan of using the machine during recovery - that isn't sensible.

Personally I use GNU ddrescue for this type of DIY recovery attempt, but it has taken time and practice to become confident with how to use it, as it can need skill to get the best results. Other people use & recommend other software (some free; some commercial). Your choice of software depends on many factors like budget, experience, expected support level etc. You can search on the forum for "clone" or "cloning" to find many previous discussions. Best to attach the "problem" disk via (S)ATA, not USB; the disk holding the target of the clone can be attached via USB if needed.

After you have cloned as much as possible from this disk, then you can attempt to recover the data from that cloned image (or, even safer, from a copy of that cloned image) - and/or you can clone that image onto a new disk. There are many different possible approaches.

Again, to be clear, there are risks with any DIY recovery attempt (I've mentioned several in previous postings), so you have to consider whether you want to take the risks. Your data; your risk; your choice.

Backup data and replace, possibly ask someone with dedicated tools. You then can ask someone in your area who REPAIR PROFESSIONALLY disks to fix it for re-use. Otherwise it is only a waste of time : backup and replace.

Thank you guys..I have recovered a few TB in the past, as favor for friends, but I do see the noobiness in me, doing the old check disk and all that, putting my data in danger of over diagnosing.
I certainly dont mean to recover data from a disk ive booted my os from, that would be silly now.
I understand that having it done professionally is always the better option, pity I cant afford it. but I happen to have access to a range of commercial data recovery software that I'm not familiar with, maybe I should look for pages on that topic while having my disk cloned by ddrescue

anyway, thanks

Yes clone it with DDresuce and then try RStudio on it. You can test it for free and if it works you can purchase this and use it to get back your data. Bad idea to run too much check disk on this one just clone it and try Rstudio and see what the results are on this one.

_________________
Iorana Haraharaini

so i tried cloning the partition using ddrescue, but it only succeeded in recovering about 36 mb of a 250 GB partition. then I used safecopy and cloned the partition into a imagefile, but couldnt mount the image using conventional methodes (loop mount, ntfs-3g etc) because appearently there is no recognizeable file system on the image file

is there any other methodes you guys would suggest? any comercial software that could clone the partition (or as much of it thats left)?
i happen to have norton ghost on windows, but windows doesnt recognize the drive at all

help?

Hi,

Now that you have a clone you want to run dr software on the image file. As mentioned before try the R-Studio demo from http://www.r-tt.com/
You will have to open the image file in R-Studio first then right click & scan the image file.
This process can take a while & will scan all the sectors in the image looking for files.
If your lucky then save your data.

If the MFT is corrupted then you will not get the folder structure & file names as these are stored in the MFT but you will get random names for the files & most important the data.

All the best

Loki

@savagerock,

Only after you received some initial replies, you then said:


... but you haven't listed the software which you have access to. Now you are asking:


It seems there is some miscommunication here. It doesn't make sense to me why you are asking us about commercial software, which you say that you already have access to... You can go to the websites of the vendors, for more details of whichever software you are talking about. However there is no information to say that commercial software would be more successful than what you have tried already.

You say that you tried ddrescue, but you don't mention the exact commands that you used, and what happened at each step. Did you just run ddrescue with its defaults? You also didn't mention how successful safecopy was.

Both ddrescue and safecopy try to run automatically with sane defaults (although they can get different results, depending on exactly how the faulty drive is behaving), but both of them can need the user to make interactive changes to the plan, depending on what happens during a recovery attempt. Creating a bad block logfile is essential with either utility, to allow recoveries to be restarted with different options. Unfortunately their logfile formats are incompatible with each other (at least they were when I last checked). That means that to avoid unnecessary repeating of reads to the failing drive, either you need to write some awk / perl / etc. to convert the logfile formats between the two utilities, depending on which one you want to try - or it may be more efficient to stay with using just one utility (which is what I would do).

As you have seen, both those utilities are very configurable. Without more details from you, starting with exactly what commands you tried (especially when using ddrescue, which I'm more familiar with), then it's impossible to make useful suggestions which may improve your result IMHO. Unfortunately with the lack of detail provided so far, I think it is becoming more likely that software-only cloning, especially without lots of experience, is not going to be very successful.

As I said earlier, after you have got the best clone that you can, there are several different possible approaches for the next steps (partly depending on how incomplete the clone is, as well as your budget, available disk space etc.) - loki has kindly described one of those approaches to you.

@Vulcan

forum's policy restricts linking to "warez" which I reckoned is any kind of copy righted material, maybe I was being too cautiose, but as we both know many of those products are available online. I did mention norton ghost, I've also got acronis true image. now that may not be exactly the power tool that you have in mind, but I figured, if there is a go-to software that can get the job done, the "HDD guru"s are the ones I should ask them about. anyway

I ran ddrescue with -d and -f parameters and 3 retries. it read around 35 mb of the partition, then encountered an error, and then just kept on couting the "time since last successfull read". as I mentioned it's a ntfs partition, which looks perfectly normal when mounted in ubuntu, but returnes I/O errors when I try to access my data, if restarted, the disk is no longer shown in bios, but with a proper shut down it appears again. the painfully long process of ddrescue somehow makes try/error approach inefficient. so please let me know the required details for determining the correct course of action.

Thanks for the info.


I didn't realise that you meant warez in your previous comment - I guessed you meant legal utilities (but we didn't know which ones) e.g. at your workplace or similar, but with the lack of earlier details, any guess could have been correct. Now it is clear what you were hinting.


Neither of those is sensible for this type of problem especially now that you have given more details than before.


That is vital new information. That behaviour is typical of some recent Seagate & Maxtor (STM) drives, which deliberately prevent further access until the next power-cycle, under some conditions. I have seen this many times in the past, in my (non-DR) work with Seagate drives. I'm sure the DR pros see this much more often than me.

Assuming your drive is behaving in the way that I have seen, then unfortunately that meant that as soon as the disk stopped returning data, all attempted reads of the remaining disk space (500GB-35MB) were a waste of time, and were not going to even be attempted by the drive, without intervention. My guess is that a weak/damaged head seems likely in your case, given your description.

With previous cases on this forum of drives with this behaviour, I have not seen successful recovery using free software-only techniques. However DR companies using hardware imagers - specifically DDI - has been more successful.

If the data has minimal value to you, and is not worth the cost of using a DR company with a DDI unit (i.e. if you really want to continue taking the risks of DIY, even if that makes any subsequent recovery by a DR company more difficult / more expensive / perhaps impossible, if you change your mind later about using these services) then you could try imaging the disk in reverse (i.e. going from high LBAs towards low LBAs). This will not be successful if the drive has a weak head, for example (in this case, again errors will start quickly and then no more data will be read, just like you have seen already), but this approach may help if the problem is due to media-related issues near the low LBAs.

Other DIY approaches are possible, and some have been tried before by other people on the forum but as I said, I have not seen any that were successful, for disks with this specific behaviour. Perhaps you will be lucky and another reader has had more success than I have seen, can suggest a successful DIY approach on a drive with those symptoms.

@Vulvan
it's a pity about the disk, nevertheless. thank you very much for your time.

I have recovered many drives in such condition using drive's own heads but had to do some "magic".
Having an HW imager at hand I would give it a go, usually it works.
Or a combination of intervention + HW imaging = successful data recovery. THEN the drive can be refurbished for re-use (non-critical data, this time!).