Czy wolisz polską wersję strony elektroda?
Nie, dziękuję Przekieruj mnie tamhow to recover data from failed hard drive
To recover data from a failed hard drive:
The most important rule: recover from a clone, not directly from the failing drive.
Hard drive failures usually fall into these categories:
| Failure type | Typical symptoms | DIY recovery? |
|---|---|---|
| Logical failure | Drive detected, partition missing, RAW file system, deleted files, accidental format | Often possible |
| Bad sectors / weak media | Drive detected but very slow, read errors, freezes during copy | Possible, but clone first |
| Electrical / PCB failure | No spin, no detection, burnt smell, visible PCB damage | Sometimes, but risky |
| Mechanical failure | Clicking, grinding, beeping, spin-up/spin-down loop | Usually professional only |
| Firmware failure | Detected as 0 bytes, wrong model, hangs BIOS/UEFI | Usually professional only |
If the drive makes repetitive clicking, knocking, scraping, or grinding sounds, stop powering it. Those symptoms may indicate head or platter damage; repeated power-on attempts can make recovery much harder or impossible.
Avoid these common mistakes:
Do not run CHKDSK, fsck, Disk Utility repair, or “repair partition” tools on the original failing drive.
CHKDSK is designed to check and repair file-system errors, but repair operations can modify metadata; use it only on a clone if data recovery is the goal. Microsoft documents CHKDSK as a tool that can fix logical disk errors when repair options such as /f are used. (learn.microsoft.com)
Do not format the drive, even if Windows says “You need to format the disk before you can use it.”
Do not initialize the disk in Windows Disk Management.
Do not install recovery software onto the failed drive.
Do not save recovered files back to the same failed drive.
Do not open the hard drive casing.
HDD platters and heads require a cleanroom environment. Dust contamination can cause additional head crashes.
Do not use the freezer method.
This old trick is unsafe for modern high-density drives because condensation and mechanical stress can worsen damage.
Connect the drive as a secondary drive to another computer if possible.
For SATA HDDs:
Check:
If it is not detected at all, or it is clicking/grinding, skip DIY software recovery and go to a professional lab.
For a failing drive, ordinary copy tools are poor because they often stop on read errors. GNU ddrescue is designed for this situation: it copies readable areas first, skips bad regions, and uses a map file so recovery can be resumed and retried efficiently. (gnu.org)
You need:
Boot Linux and identify disks:
lsblk -o NAME,SIZE,MODEL,SERIALBe extremely careful identifying the failed source drive and the healthy destination drive.
Example:
/dev/sdX/dev/sdYrescue.mapFirst pass: copy good sectors quickly, skip bad areas:
sudo ddrescue -f -n /dev/sdX /dev/sdY rescue.mapSecond pass: retry bad areas:
sudo ddrescue -f -d -r3 /dev/sdX /dev/sdY rescue.mapImportant:
/dev/sdX = failed source drive./dev/sdY = healthy destination drive.rescue.map file; it lets ddrescue resume without starting over.You can also clone to an image file:
sudo ddrescue -n /dev/sdX /mnt/recovery/failed_drive.img /mnt/recovery/rescue.map
sudo ddrescue -d -r3 /dev/sdX /mnt/recovery/failed_drive.img /mnt/recovery/rescue.mapAfter cloning, disconnect the original failed drive and work only on the clone or image.
Once you have a clone, choose the correct recovery method.
Use TestDisk.
TestDisk is intended for recovering lost partitions and repairing boot/partition structures. Its companion tool PhotoRec is used when file-system structure is too damaged and files must be recovered by signature scanning. (photorec.dev)
General TestDisk workflow:
Do not write repaired partition structures to the original failed disk.
Use PhotoRec or a commercial file recovery tool.
PhotoRec performs file carving: it scans raw sectors for known file signatures. This can recover files even when the file system is corrupted, but it often loses original filenames and folder structure. The official documentation notes that PhotoRec can recover files from corrupted file systems, but not necessarily original names or directory structure. (cgsecurity.org)
Use PhotoRec when:
Recover files to a different physical drive, never to the clone being scanned.
If the drive is healthy and the issue is deletion/formatting:
The less the disk has been used since deletion/formatting, the better the chance of recovery.
Likely causes:
Action:
Possible causes:
This is often beyond normal software tools. Professional labs use hardware imagers and vendor-specific repair systems to stabilize the drive before imaging.
Possible causes:
A technically skilled person may inspect the PCB, but PCB swaps are not simple on modern drives. Many HDDs store unique calibration/adaptive data in ROM or flash on the PCB. A donor PCB often requires ROM transfer or firmware adaptation.
If the data is valuable, professional recovery is safer.
If the “hard drive” is actually an SSD, the approach changes.
SSD failures often involve:
With SSDs, chip-off recovery is much harder than with older flash devices because data may be encrypted, wear-leveled, compressed, and interleaved across NAND packages. If an SSD is not detected or shows an incorrect capacity, professional recovery is usually the best option.
Also, deleted files on SSDs may be unrecoverable if TRIM has erased the blocks.
The current best-practice workflow remains:
GNU ddrescue remains one of the standard tools for unstable media because its map-file-based workflow avoids repeatedly rereading known-good areas and allows interrupted recovery sessions to resume efficiently. (gnu.org) TestDisk and PhotoRec remain widely used open-source tools for partition recovery and file carving. (photorec.dev)
Use this sequence:
Ideally have:
Example:
Failed HDD ---> Clone HDD/Image ---> Recovery software ---> Recovered files on third diskStop and use a professional recovery service if:
DIY attempts can reduce the odds of professional recovery if the problem is mechanical.
Only recover data from drives you own or are authorized to access. If the drive contains business, medical, legal, or personal data belonging to someone else, preserve chain of custody and avoid altering the original media. In professional environments, image the drive first and document every action.
For a failed hard drive, the safest approach is:
If you tell me the exact symptoms — HDD or SSD, internal or external, detected or not, any noises, operating system, and what Disk Management/BIOS shows — I can give you a more specific recovery plan.