bios - What damage can vibration cause on hard-drive?
2014-06
Pieter
My laptop's hard-drive completely died last week. It started with a boot-up failure, so I started up a live USB to inspect the drive. This didn't really work as it seemed to freeze the system up, but I did get an error about a fault in one of the sectors/partition that held the OS.
Because of the freeze ups, I restarted into the live USB to try and repair the drive, but it was no longer picked-up and BIOS doesn't see it any more. I putted it in another laptop and its BIOS doesn't see it either.
So now I'm trying to recover some data since the current back-up is two months old (forgot to set it up after distribution upgrade) and came across this question.
I know that a 'propeller' broke on the fan and caused a bit of vibration when it was on and suspect that it might have caused the damage. So I want to know what kind of damage can such vibration cause so that I can consider the recovery options?
Matthew Williams
Hard drives can be very resilient to lots of different things, but vibration can damage them. Your hard drive will have mechanical movement as heads move very quickly (7200 RPM) at very close proximity to the platters, which store your data. Vibration can cause the heads to connect with the platters at times they aren't supposed to, kind if like the needle of a record player touching a record where it isn't supposed to.
In addition to the above issue, vibrations can cause the heads to become unstable making data reading very difficult, but from your description you seem to have the former problem (if vibration is the cause).
In terms of recovery you could put the drive in a caddy and run TestDisk to see if anything can be done, but if your drive is physically damaged I think this might be a hard lesson in the use of good backups.
Little Helper
I had a damaged disk, I didn't know that. But after some use, other problems started to appeared. So I just replaced it because of it failures.
So what are the most common causes of Block failure? And why hardware can't ignore the damaged Block (sometimes applications crashes, BIOS can't boot from that disk, or Windows cannot properly install)?
CJM
The two main causes of damage (whether that be at block level or anything more significant) are:
- Manufacturing errors - the problem may be there from the outset or may appear after some time/use
- Impact damage - disks can only withstand shocks up to a certain tolerance before damage is likely to occur.
If you have one or more bad block, the disk will usually detect this and try to recover information which will be moved to one or more spare blocks. The original block will be marked as 'bad' and will no longer be used. Of course, depending on the problem, this data recovery may not be possible - hence the problems you describe. Equally, there are a finite number of spare blocks, and once these have been filled, your problems will accelerate.
While you can sometimes live a small number of bad blocks, it is usually the start of a slippery slope, so it is best seek a replacement disk - if your disk is within it's warranty period, you may find you can get a free replacement - I have successfully RMA'd several disks because of bad blocks.
Most modern disks support SMART - which can give warnings of disk hardware problems, and there are plenty of tool, such as HD Tune what can check SMART information, run surface checks (for bad blocks) and measure performance.
gd1
Making it very simple and kind of inaccurate, an hard disk is basically a set of one or more magnetic rotating platters which are read by a head that is managed by a controller (not the SATA controller of your mobo, a controller into the disk itself).
I don't know all the possible reasons why a block of a magnetic platter can become 'damaged'. I heard that, when the disk suffers some physical stress, the head can crash on the platter (they are very close!) damaging the incredibly fragile magnetic film which is on the platter itself.
Also, a defective manufacturing of the magnetic platter can soon bring to read failures.
A 'damaged' sector either cannot be read at all, or at least, it can be read but with frequent read errors -- so that it is considered 'unreliable'. The disk controller (if it works -- and sometimes it can be damaged or defective itself) can recognize sectors which are impossible or hard to read/write and mark them as 'damaged', so that they won't be read or written anymore. Modern disks recognize and isolate bad sectors without you knowing and often without you loosing any data. So I'm a bit surprised you had so many problems. Maybe the disk was very old so that it didn't implement proper bad sectors handling, or the controller was defective, or the head itself was screwed, or there were so many bad sectors that using the disk became impossible.
However I'm fine and happy with a disk of mine which has 15 bad sectors. I do backups often, obviously.