Features
Backup drive failure after a drop: what actually survives long-term storage
How a dropped backup drive fails, what hard drives and SSDs survive, and how to test, image, and replace a suspect drive for long-term storage.
What to take away
- A dropped hard drive often works for hours and then fails, because the heads spread surface damage as they pass over it.
- A drive switched off during the fall usually loses its enclosure, port, or circuit board, not its platters.
- Solid state drives tolerate shock better, but unpowered retention shortens in warm storage.
- Image any suspect drive with ddrescue before running recovery software, and never write to the original.
- Verify backups by checksum, and keep three copies with one offsite.
What a drop physically damages
A hard disk drive writes with heads that float above spinning platters on a thin cushion of air. The flying height is measured in nanometers. A shock while the drive is powered can push a head into the recording surface, and the scratch destroys sectors no software can read.
The hard disk drive article covers that mechanical fragility, including the shock ratings vendors publish for powered and unpowered states.
An unpowered 2.5 inch portable drive is a gentler case. Most models park their heads when power is cut, so a fall may only crack the shell, bend the USB connector, or loosen the bridge board.
The circuit board deserves a look. A drop can crack solder joints or flex the board enough to break a trace. On most modern drives the board holds calibration data tied to that unit, so a donor board swap usually fails.
Example: a two-meter drop onto a wooden floor
Consider a 4 TB portable drive that falls off a desk while it is copying files. It remounts, copies for twenty minutes, then vanishes from the file manager. That sequence points to head or platter damage, because the drive reads until it reaches the damaged zone.
Two-meter drop onto wooden floor
| What you see | Likely cause | First move |
|---|---|---|
| Clicking, then the drive disappears | Head or platter damage | Power down, do not retry |
| Mounts, but a folder is empty | Damaged sectors in the file table | Image the drive, then scan the image |
| Not detected on any computer | USB bridge, port, or board damage | Try the bare drive in another enclosure |
| Beeps or spins without mounting | Seized spindle or firmware fault | Stop and contact a cleanroom lab |
Failure signs to check before you trust the drive again
- SMART attributes, above all reallocated sectors, pending sectors, and uncorrectable errors.
- A full surface scan from the maker's tool, such as SeaTools or Western Digital Dashboard.
- The sound of the drive at idle and during one long copy.
- A read-back of every file against a checksum stored when the backup was written.
Many USB enclosures do not pass SMART data to the host. Use a SATA port, or an enclosure that supports UASP. Free tools such as smartmontools and CrystalDiskInfo read the same counters, and Backblaze publishes annual drive stats on real failure rates.
Long-term storage: what survives years on a shelf
An unpowered solid state drive slowly loses the charge that holds each cell's state. JEDEC standards define retention for powered-off storage and tie it to temperature, so a drive kept in a hot garage ages faster than one in a cool closet. The solid state drive article covers shock resistance and retention in more detail.
Hard drives age differently. Magnetic media can hold data for years, but the mechanical assembly is the weak point. Lubricant migrates, spindles seize, and a drive left on a shelf for five years may not spin up.
Bit rot affects both types. Files can decay while the drive still reports healthy. Checksumming file systems such as ZFS and Btrfs catch this, and par2 files repair damaged archives elsewhere.
A backup you have never restored is a guess, not a backup.
NIST SP 800-88 explains how to verify and sanitize media before you reuse or retire it, which matters when an old drive still holds tax records.
Recovering data after impact: a workable order of operations
Recovering data after impact
- Power the drive down and leave it down. Each extra spin-up can widen existing platter damage.
- Test the cheap parts firstcable, port, power adapter, and a second enclosure.
- Clone the drive with ddrescue onto a larger healthy target, using its log file so an interrupted run resumes.
- Run file recovery against the clone only, never against the original.
- If the drive clicks, beeps, or is never detected, send it to a lab with a cleanroom.
Cleanroom work is priced per case, so ask for a written quote and a no-data, no-fee term. A cracked connector or failed bridge board is far cheaper to fix than scratched platters.
Deciding whether the drive is worth replacing
After a serious impact, a drive becomes a liability. Use it for scratch data if a full surface scan comes back clean, and keep nothing on it that you cannot afford to lose.
A safer replacement plan is two drives rather than one, plus the 3-2-1 rule: three copies, two kinds of media, one copy offsite. If you are buying rather than recovering, our External SSD vs HDD durability comparison covers drop ratings, unpowered retention, and cost per terabyte: External SSD vs HDD durability.
