1 |
On 2/1/06, Grant <emailgrant@×××××.com> wrote: |
2 |
> Thanks Peter. That is quite contrary to what most of the other posts |
3 |
> in this thread are saying. Those are all just rumors and myths? |
4 |
|
5 |
I think for what would be available for you, me, or ${megacorp} to |
6 |
use, yes, it is rumor and myth. As I mentioned previously, the |
7 |
density of data on modern drives makes surface analysis (by which I |
8 |
mean anything that does not simply read the drive with standard drive |
9 |
electronics and and search the resulting data) very difficult. I |
10 |
would say impossible, but we simply don't know what techniques are |
11 |
available to the NSA or other government agencies to use. |
12 |
|
13 |
Again, my *guess* is that with a *very* modern drive where the |
14 |
manufacturers simply cannot squeeze any more data onto the platter, |
15 |
that even the NSA would not be able to recover any data. But it may |
16 |
be that is just what they /want/ us to think... |
17 |
|
18 |
I posted this before, but it is the best and most thorough study I |
19 |
could find on this topic: |
20 |
|
21 |
http://www.simson.net/clips/academic/2003.IEEE.DiskDriveForensics.pdf |
22 |
|
23 |
And another paper referenced in the above study: |
24 |
|
25 |
http://www.cryptoapps.com/~peter/usenix01.pdf |
26 |
|
27 |
This paper talks about physical scanning of memory devices for |
28 |
encryption keys, so is a bit off-topic (even for this off-topic |
29 |
thread), but it should give you a hint of what kind of effort would be |
30 |
required to try and recover overwritten data from a hard disk. |
31 |
|
32 |
Just a quick quote from the Peter Gutmann paper: |
33 |
|
34 |
---- |
35 |
Finally, however, the best defence against data remanence problems in |
36 |
semiconductor memory is, as with the related problem of data stored on |
37 |
magnetic media, the fact that ever-shrinking device dimensions (DRAM |
38 |
density is increasing by 50% per year [74]), and the use of novel |
39 |
techniques such as multilevel storage (which is being used in flash |
40 |
memory and may eventually make an appearance in DRAM as well [75]) is |
41 |
making it more and more difficult |
42 |
to recover data from devices. As the 1996 paper suggested for magnetic |
43 |
media, the easiest way to make the task of recovering data difficult |
44 |
is to use the newest, highest-density (and by extension most exotic) |
45 |
storage devices available. |
46 |
---- |
47 |
|
48 |
-Richard |
49 |
|
50 |
-- |
51 |
gentoo-user@g.o mailing list |