Destroying a hard drive
Ben Scott
dragonhawk at gmail.com
Wed Jul 11 15:41:00 EDT 2007
On 7/11/07, Tom Buskey <tom at buskey.name> wrote:
>>> FWIW, that is not considered sufficient for media containing
>>> classified information.
>
> I think we melt them.
That'll do it. Our DSS IS rep stated that acceptable methods for
destruction of magnetic media are:
- Degauss with something from the NSA EPL
- Grind to fine powder
- Burn to fine ash
- Phase change (liquefy or vaporize)
> I know paper gets burned.
Requirements for paper are a lot less stringent than computer. An
NSA EPL cross-cut shredder is sufficient. (The shreds are tiny, like
1 mm square.) I presume the difference is information density. You
can't fit much information on human-readable paper. But even a tiny
fragment of a disk could have hundreds of pages worth of data.
> You want to make the cost of recovery more expensive then
> the worth of the data to whoever might get the drive.
*Exactly*.
But it is important to be aware of the state-of-the-art in
commercial data recovery. There are many commercial services which
specialize in this stuff, and they have equipment like clean rooms,
spare parts, and specialty readers. Failures in electronics or motors
are almost trivial for them to fix. Bent platters may be almost 100%
recoverable. Holes are good, but anything not drilled out may be
recoverable. Prices have been between $1000 and $1500 for the stuff
I've had to deal with.
In many industries, someone conducting industrial espionage might
find $1500 a worthwhile investment.
On 7/11/07, VirginSnow at vfemail.net <VirginSnow at vfemail.net> wrote:
> I wonder how a drive would fare if you opened it up, stuck it in a
> microwave oven, and put it on High for a couple of minutes...
I dunno if that would effect the magnetic properties or not.
(Secondary effects might do it, such as when your house burns down
from the resulting fire...)
On 7/11/07, Jim Kuzdrall <gnhlug at intrel.com> wrote:
> How much contiguous track is needed to record a 4KB file?
Let's see. Wikipedia gives 150 Gbit/in². 4 KB = 32768 bits. 150
billion divided by 32,768 equals 4,577,636. One five-millionth of a
square inch?
> Drilling is a good cheap alternative, and most likely effective
> enough for my present purposes.
For my own stuff and regular business data, I use DBAN, followed by
drilling several holes through through the drive (including platters).
If DBAN can't work because the drive is too far gone, I drill more
holes then otherwise. :)
> The problem with magnetic erasure is that I have no facility to
> confirm that it was actually successful.
That's why you use an NSA EPL degausser, rather than a fridge
magnet. FYI, degaussing destroys the factory low-level formatting,
supposedly rendering modern drives unusable anyway.
> My wife suggested a magnetic erasure that might be more economical
> than the shredding truck. She suggested renting the MRI machine at the
> hospital.
Magnetics is a funky science. I'm not sure that'll do it.
Obviously an MRI is a big magnet, but does it have the right... I
dunno... "density", for lack of a better word?
Either way, I kinda doubt a hospital would want to risk their
expensive MRI machine being damaged by a hard disk flying up against
it. :)
> So far I like the idea of melting the drive best. Perhaps someone
> could put a drive on the charcoal grill at the gnhlug BBQ. Is it hot
> enough?
For most grills, that would melt/burn the plastic, but I don't think
it would effect the metals much. On the other hand, there's this
grilling method:
http://www.ambrosiasw.com/Ambrosia_Times/September_95/2.5HowTo.html
-- Ben
More information about the gnhlug-discuss
mailing list