1ComputerCare   Bookmark and Share
     
   
     
  Bad Blocks  
     

The single most common hardware fault I deal with relates to failed disk drives; apart from cooling fans these are normally the only components that have moving parts.  A disk drive comprises a controller card that contains electronics to manage the data access and the mechanical part, this is basically a collection of platters which rotate with read/write heads hovering a few nanometres above and below.  These are what hold the files and the parts that most commonly fail.

Why do they fail ?

  • This could be due to a virus or other program (e.g. low level editor) that has been set - accidentally or deliberately, to mess with the file system.

  • If Windows was stopped abruptly as in a power failure or the on/off button being pressed for more than 4 seconds, then any files held in volatile memory would not be written to disk thereby causing a corruption next time the computer is started (this is a soft fault).

  • It could be that the disk has been dropped or received a sudden jolt causing the read/write head to come into contact with the fragile platter on which the data resides.

  • Maybe excess heat (from a failed heatsink/fan or a close radiator) or a magnetic force (from a speaker perhaps) caused a disturbance.

  • Electrical distortions can cause this too - make sure an active surge protector or uninterruptible power supply is shielding the disk from these problems.

  • Most commonly I suspect, a manufacturing process flaw causes it. Perhaps the manufacturing process was not as clean as it could be or the structure of the platter was sub standard.

What is a bad block, cluster or sector ?

The disk platter is logically broken into small units to enable the data to be passed efficiently.  Typically, data is held in bits which hold a value of positive or negative polarity.  There are usually 8 bits in a byte, 512 bytes in a sector (block is synonymous with sector) and 4 sectors in a cluster.  These values vary with the file system in use but are typical.

Remapping.

Up to a point, bad blocks are a normal function of a disk and the manufacturer will allow a hidden region of the disk to house pointers to known bad blocks.  By mapping the blocks to this hidden region, Windows can not allocate the block and it therefore causes no further problem.  Problems do occur though when the number of bad blocks exceeds the capacity of the bad block region to house them.  The remapping of bad blocks is often called 'repairing' but this is rather a misnomer as the blocks once damaged, are simply hidden from normal operation.

How to avoid or mitigate problems.

  • Use a sure protector or UPS.
  • Run a disk monitoring program in the background e.g. Acronis
  • Do not power off abruptly.
  • If the computer crashed with a blue screen, unplug the computer from the mains until after professional help has been sought.
  • Take backup copies of files frequently and then verify that you can recover them when necessary.
A single bad block event is an inconvenience, but if it becomes one of a series, you know that the disk is on the way out and it would be prudent to [have me] clone the disk to a new one while it is still possible.

Please let me know how this page could be improved.