[wellylug] wellylug Digest, Vol 58, Issue 2

[C.T.F. Jansen]

Greetings,
          Regarding making previously written data on a hard drive
unreadable and the unreliablility of very expensive methods to extract
data from a drive that has been over written.
  I've watched engineers trying to extract whatever they could from a
large old hard disk with some equipment they'd brought with them. The
drive happened to have suffered a disk crash with me the customer
literally breathing down their necks.
  Despite the glittering array of equipment intended to recover whatever
could possibly be gotten off the disk there was no more data forthcoming.
The terrible spectre of data loss from yet another in a string of
disk crashes hung over their efforts of some hours. There wasn't any more
data loss while they were busy. They told me the factory couldn't recover
any more, perhaps they were right.

  With an alternating pattern of 1 and 0 throughout the disk done a number
of times the magnetic field patterns around the bits will render any
residual field patterns indistinguishable from noise. I would be very
interested in any studies done that have successfully extracted any 
previously written data from a disk randomly over written a number of
times.
  Speculation is one thing but actual data recovery practise is another.
When a database catalogue has been overwritten how often is the rather
valued data recovered from the disk where it was ? Ever ?

  Magnetic tape is a different matter because its physically bigger,
wider, with a lot more slop.

Cheers,

frank.jansen at actrix.gen.nz