[swinog] Re: How to destroy data effectively?

2022-12-07 Diskussionsfäden Mark Walliser via swinog

I am a big fan of encryption and in particular FDE.

But I still tend to disagree on the approach using FDE to "destroy" no 
longer used data on physical media. There are just too many variables 
beyond ones control. Just to name a few:
- Will the encryption algorithm still be secure in 10, 30, 50, ?? years? 
(Old data may still be valuable data.)
- Is the integration secure? (How do you make sure, that the data on the 
platters - or in case of SSDs in the memory cells - really IS encrypted 
the way you expect it?)

- Possible backdoors in the algorithm?
- Possible backdoors in the integration?
- Are the "lost keys" really lost? (Might depend on the integration 
again and of course the sysadmin.)


I personally see it the other way round: physical destruction is the way 
to go - FDE is bonus (e.g. for the "my disk got stolen" case).
If the data really is high profile, the "disk broke" case should be 
handled with physical destruction and a new disk should be bought.


Greetz,
Mark

Am 04.12.2022 um 13:47 schrieb Jeroen Massar via swinog:

The real answer, net to using it for target practice, shredding and melting 
down is much easier: Full Disk Encryption.

Just lose the encryption keys and the data is useless. If you then also do one 
of the above for fun, just added bonus.

FDE helps for the "my disks got stolen" case, but also for the "disk broke" 
case, and just letting a random remote hands person remove them: one does not have to trust that 
they are destroyed properly, as nobody, but hopefully the sysadmins, have the FDE keys.

Of course, FDE does not help when the disk is online and one can SSH or 
otherwise execute code on it, but that is a different problem.

Regards,
  Jeroen

PS: Food for thought: what is worse, Financial Services or Advertising?
 [and at least you are not scamming people with ponzi schemes, right...? :) 
]


On 2 Dec 2022, at 15:51, Martin Ebnoether via swinog  
wrote:

Hi all.

As some of you know, I work at a money laund... financial
company. Some time ago, the question arose, how to effectively
destroy data safely and securely in an easy way?

How does your company deal with hard disks (or any media) that
needs to be decommissioned? Do you just dd a few times over it?
Or rather let a professional company shred your media to little
bits?

CU, Venty

--
10 PRINT "BASIC programmers don't die."
20 PRINT "They just GOSUB without RETURN."
30 END
READY.
___
swinog mailing list -- swinog@lists.swinog.ch
To unsubscribe send an email to swinog-le...@lists.swinog.ch

___
swinog mailing list -- swinog@lists.swinog.ch
To unsubscribe send an email to swinog-le...@lists.swinog.ch


___
swinog mailing list -- swinog@lists.swinog.ch
To unsubscribe send an email to swinog-le...@lists.swinog.ch


[swinog] Re: How to destroy data effectively?

2022-12-07 Diskussionsfäden Jeroen Massar via swinog


> On 7 Dec 2022, at 13:04, Hendrik Jäger via swinog  
> wrote:
> 
> Hi
> 
>> And this is a problem if you rely on something you can not verify 
>> immediately. For example if I use a big hammer I immedialtey  see the 
>> results. But a degaußed Disk does not looked destroyed - you can not verify 
>> it with your eyes.
> 
> You see the physical result but does that really reliably mean that the data 
> is not recoverable?
> I’m just thinking of the work some germans are doing to reconstruct 
> shreddered stasi files: they also seemed completely destroyed, at least 
> enough that the stasi considered it enough, yet they are being reconstructed.
> I’d imagine that a hammer would not be enough to be _certain_ that 
> reconstruction is _impossible_ (not just more or less convinced that no one 
> will put in the effort to attempt it). Are you sure it is enough? When is it 
> enough? I imagine bent platters are hard but not quite impossible to 
> reconstruct and the effort required would probably not be worth the results 
> in most cases. But that always depends on the significance of the data on the 
> disks …
> I wouldn’t feel _certain_ with neither hammer nor degausser because I’m not a 
> recovery expert. Melting the platters down with just heat or thermite or 
> something would probably convince me. Shredding them to 1x1mm tiny pieces 
> would leave me reasonably certain enough for most scenarios, as well.
> 
> Any data recovery experts on this list who can shed more light?

As I noted: Full Disk Encryption.

Throw away the encryption keys (forget them) and you are done.

It solves the "disks get stolen" and the "we need to destroy the disk" part, 
noting that when a disk fails you cannot write to it anymore.

Any physical destruction is then just for show.

Greets,
 Jeroen

___
swinog mailing list -- swinog@lists.swinog.ch
To unsubscribe send an email to swinog-le...@lists.swinog.ch


[swinog] Re: How to destroy data effectively?

2022-12-07 Diskussionsfäden Hendrik Jäger via swinog
Hi

> And this is a problem if you rely on something you can not verify 
> immediately. For example if I use a big hammer I immedialtey  see the 
> results. But a degaußed Disk does not looked destroyed - you can not verify 
> it with your eyes.

You see the physical result but does that really reliably mean that the data is 
not recoverable?
I’m just thinking of the work some germans are doing to reconstruct shreddered 
stasi files: they also seemed completely destroyed, at least enough that the 
stasi considered it enough, yet they are being reconstructed.
I’d imagine that a hammer would not be enough to be _certain_ that 
reconstruction is _impossible_ (not just more or less convinced that no one 
will put in the effort to attempt it). Are you sure it is enough? When is it 
enough? I imagine bent platters are hard but not quite impossible to 
reconstruct and the effort required would probably not be worth the results in 
most cases. But that always depends on the significance of the data on the 
disks …
I wouldn’t feel _certain_ with neither hammer nor degausser because I’m not a 
recovery expert. Melting the platters down with just heat or thermite or 
something would probably convince me. Shredding them to 1x1mm tiny pieces would 
leave me reasonably certain enough for most scenarios, as well.

Any data recovery experts on this list who can shed more light?

Thanks

Hendrik

On Wed, 7 Dec 2022 08:25:01 +0100
Klaus Darilion via swinog  wrote:

> > -Ursprüngliche Nachricht-
> > Von: Per Jessen via swinog 
> > Gesendet: Dienstag, 6. Dezember 2022 22:40
> > An: swinog@lists.swinog.ch
> > Betreff: [swinog] Re: How to destroy data effectively?
> > 
> > Martin Ebnoether via swinog wrote:
> >   
> > > Hi all.
> > >
> > > As some of you know, I work at a money laund... financial
> > > company. Some time ago, the question arose, how to effectively
> > > destroy data safely and securely in an easy way?  
> > 
> > When I worked in money laun... finance myself, in the 1980s, we used
> > _large_ electromagnets.  
> 
> We used to use a "Degaußer" for hard disks. Recently, a colleague put the 
> degaußed disk back into his PC and could read data without problems. So, 
> either the Degaußer is broken or for whatever reasons it did not work. And 
> this is a problem if you rely on something you can not verify immediately. 
> For example if I use a big hammer I immedialtey  see the results. But a 
> degaußed Disk does not looked destroyed - you can not verify it with your 
> eyes.
> 
> So, we do not rely on degaußing anymore without verifiying (reconnect and 
> test). So we prefer the hammer or commercial services (which still can fail, 
> but then you can sue someone)
> 
> regards
> Klaus
> ___
> swinog mailing list -- swinog@lists.swinog.ch
> To unsubscribe send an email to swinog-le...@lists.swinog.ch



-- 
XMPP: h...@netwichtig.de
IRC: irc://irc.linux-geeks.org/
Web: https://henk.hnjs.ch
___
swinog mailing list -- swinog@lists.swinog.ch
To unsubscribe send an email to swinog-le...@lists.swinog.ch