[PLUG] Data Recovery Services in Portland

Ronald Chmara ron at Opus1.COM
Wed Jun 7 07:05:46 UTC 2006


On Jun 6, 2006, at 11:19 PM, Larry Brigman wrote:
> On 6/6/06, Auke Kok <sofar at foo-projects.org> wrote:
>> Russ Johnson wrote:
>> > Steve Bonds wrote:
>> >> Although I've heard rumors that such levels of data recovery are
>> >> possible (e.g. after a disk wipe) I've never seen it actually done.
>> >> Have you?
>> > Yes. I used to work just outside the data recovery lab at Symantec. 
>> I've
>> > also seen floppy disks reassembled after being cut with scissors.
>> > I've even recovered data off hard drives, with shareware tools from
>> > download.com, after the Windows 98 drive was nuked and paved with 
>> linux.
>> >> For all practical purposes one pass of writing zeroes is enough to
>> >> thwart a data recovery.
>> > Obviously you've never realized why the DOD requires 7 passes of a 
>> data
>> > scrambler for "secret" data, and more for "top secret".
>> and as a matter of fact, writing all zeroes (or any regular pattern) 
>> is making
>> it *very* easy to recover the data ;^)
> It has been almost a decade since the actual data save on the disk
> resembles the data
> you want.  It all gets encoded to ensure that you have a 'non-return
> to zero' pattern on the
> disk.

Good thing none of us are still using or maintaining any hardware and 
software more than a decade old.... I mean, mainframes just get 
replaced every 3 years, just like laptops, right? And nobody would ever 
have the time, or money, to reverse engineer out how to "read" the 
altered data patterns?

In my previous gig this year (!) I was writing snort-compatible IDS 
script algo's for circa 1992 (IIRC) DG/UX 5.4.1 88K *minis* that run 
inside some class B scale networks... the old boxen die much harder 
than the new cheapness, and if a piece of data is worth, oh, several 
million (or billion) dollars, it's not a good idea to underestimate the 
bizarre extremes people would take to reconstruct it.

:-D

Personally, I find a good 7 passes over the platters with a 
sledge-hammer to be quite effective, though safety goggles are a must. 
One can induce randomness of zeros and ones by flipping over the 
platters every pass with the sledgehammer, or, for the extremely 
paranoid, a *magnetized* sledgehammer head might be a good toy.

A sledgehammer also looks pretty darned cool in a server room, but if 
working with buggy OS's, may need to be placed inside a "break glass 
only in case of extreme server frustration" case.

-Bop
--
4245 NE Alberta Ct.
Portland, OR 97218
503-282-1370




More information about the PLUG mailing list