Re: destorying the hard drive
There is only one controlled and verifiable experiment that I know of
where data was "apparently" recovered on a wiped drive, this is the so
called "HelpHelp" test conducted by researchers at the Center for
Magnetic Recording Research at the University of California. The test
was conducted to determine the efficacy of Secure Data Erasure.
Before anyone makes assumption that this proves that data recovery is
possible on securely wiped drives they should read the papers that
detail how the test was conducted. In particular they should note the
following "skewed" factors which gave the researchers every possible
advantage in their search for the overwritten data, their mission was to
defeat secure data erasure and to help them all the odds were purposely
stacked in their favour:
1- The researchers had to know what to look for and where to look for
it, in the case of this test they were specifically told to look for the
written bits "helphelp" on a certain track on the disk. Without knowing
what to look for and where to look for it the researcher didn't have a
clue as to what they were looking at or seeing with their fancy tools!
2- The disk track where the helphelp bits were written was completely
wiped and tested before hand, this ensured a clean track and that the
only written bits on the track before the second test wipe were
"HelpHelp". As if condition 1 above wasn't skewed enough in favour of
the test, there were no other possible bits to weed through other than
the helphelp bits!
3- The researchers had to know the overwriting pattern of the wiping
software, when the overwriting was random and unknown the researchers
couldn't find the "helphelp" bits.
4- When two wipes were done on the drive the researchers couldn't find
the "helphelp" bits, the one pass wipe done for the test was hardly a
secure wipe!
What it boils down to is that in the end the test proved conclusively
that data recovery was impossible on securely wiped drives, the
effectiveness Secure Erasure was confirmed beyond the shadow of a doubt.
Here is one paper which mentions the so-called "HelpHelp" test:
Secure Erase of Disk Drive Data (PDF, 294kb)
Gordon Hughes and Tom Coughlin
http://www.tomcoughlin.com/Techpapers/Secure Erase Article for IDEMA, 042502.pdf
John
Bill in Co. wrote:
> I certainly would expect that writing pseudorandom, or even identical, bytes
> to each and every sector on the disk would make it nigh impossible to
> recover anything, - IF that laboriously slow procedure was invoked. How it
> could possibly be otherwise makes little sense to me - unless we operate
> under the assumption that the electromagnetic writes are somewhat incomplete
> (that is, the magnetic domains on the disk are not fully reversed (or
> realigned) completely, but still have some very small residual leftover
> effects (i.e. retentivity) from a previous write operation). (I'm an EE,
> but I'm just making some basic assumptions here!).
>
>
>>John John wrote:
>>
>>>Mike Y wrote:
>>>
>>>
>>>>>you will not be able to recover any data whatsoever on a wiped hard
>>>>>drive, it can't be done, period!
>>>>>
>>>>>John
>>>>>
>>>>>And by the way, I do not work for any company that is in anyway
>>>>>involved
>>>>>in the sale or development of anything to do with computer technology
>>>>>or
>>>>>software.
>>>>
>>>>
>>>>Well, I do, and did. I've been involved with the 380 chip set way back,
>>>>and
>>>>I do know a bit about how it and hard drives in general work. And while
>>>>I've
>>>>not personally done it, I AM aware of the technologies involved, the
>>>>theory
>>>>behind the technologies, and the practice implementing those
>>>>technologies.
>>>>Granted, I've not heard much about the techniques since drives moved
>>>>into
>>>>the ZBR (a LOT changed when drives went that route) world with the high
>>>>speed transfers (compared to early MFM), but there's nothing in the
>>>>techniques or theory that would make it impossible other than that the
>>>>tools and techniques have to stay 'ahead of the game' the same way they
>>>>were then.
>>>>
>>>>It's doable. Period!
>>>
>>>It cannot be done! Period! It is a theory only and it has never been
>>>proven! Not too long ago the US Department of Defense issued a tender
>>>call for someone to provide methods to recover data from wiped drives
>>>and no one stepped up to the plate to fill the tender request.
>>
>><snip>
>
>
>