Reformat and secure data deletion

[Golden]

it is well worth the extra time to do a DoD wipe, just to be sure. I do them when I sell a disc online or through eBay to someone I don't know. I use a zero-only option when I give a drive to a friend or family member. When I'm dealing with releasing my drives to the wild, I prefer to know for sure that my data is gone. If it takes an extra hour or so, I'm fine with that.

I agree with this approach.

Todays opportunistic criminal looking for financial gain is far more sophisicated than the previous generation. Previously they would be prepared to dig through our trash for paper documentation, today they are happier doing it from the comfort of their workstation running algorithms on a disk they found or were sold. Considering the vast amount of financial information we store on our disks, it seems prudent to be a little bit more careful.

Safe computing,
Golden

 

[FliGi7]

All data is unrecoverable after one pass of 0's to the full drive. Doing more passes than one is superfluous.

If seen plenty of people suggesting otherwise. I don't know enough to say which is right, but considering what a person has to lose, it is well worth the extra time to do a DoD wipe, just to be sure. I do them when I sell a disc online or through eBay to someone I don't know. I use a zero-only option when I give a drive to a friend or family member. When I'm dealing with releasing my drives to the wild, I prefer to know for sure that my data is gone. If it takes an extra hour or so, I'm fine with that.

I happen to know about this in particular and one pass will render all data unrecoverable. Even an STM, as Monsoon mentions, will not be able to recover it. Any empirical tests that have been done have only been able to "recover" data that they knew was there before-hand (i.e. they were able to deduce from circumstantial evidence that a 1 was there prior to the wipe by knowing it was there before-hand). The reasoning behind any testing for recoverable data boils down to the fact that 1's are never 1.00000 and zero's are never 0.00000 on disk.

At any rate, there has never been a case of substantial (more than finding a possibly conclusive remnant one or zero) data recovery via STM after even one pass of zero's. Two or more passes is for the mental benefit, not actual.

 

[Golden]

Hi Fligi7,

If that is the case, what are the origins and reasoning behind such methods as the Gutmann 35 Pass method?

Regards,
Golden

 

[ignatzatsonic]

Hi Fligi7,

If that is the case, what are the origins and reasoning behind such methods as the Gutmann 35 Pass method?

Regards,
Golden

My guess would be paranoia. It's got a lot to answer for.

 

[FliGi7]

Hi Fligi7,

If that is the case, what are the origins and reasoning behind such methods as the Gutmann 35 Pass method?

Regards,
Golden

That was a technology developed in 1996 on older, much less dense drives where variances in permittivity were much more evident and plausible for data recovery. Soon thereafter as time progressed and hard drives became more and more dense, these nuances were essentially rendered obsolete and it was deduced that 35 passes never needed to be done, rather a couple (at most) random data passes would provide the same data recovery safety.

 

[Golden]

Thanks Fligi7. So would it be true to say that a lot of the software that contain multiple algorithms like DoD 5520, Schneider 7 Pass, Gutmann etc. are/may be obsolete?

If I understand you correctly, 2 writes (1 x 0, 1 x random?) are sufficient on modern day hard disks?

Regards,
Golden

 

[FliGi7]

They're not obsolete methods, just overkill for the goal of secure data deletion in 99.99% of cases.

For all intents and purposes of the consumer getting rid of a hard drive or selling it, one write of 0's is sufficient enough to prevent data recovery. It might be possible for someone with a Scanning Tunneling Microscope to recover bits and pieces of the data if they have the time and resources. This somewhat empirical, although not very productive or substantial, ability for recovery is what drives entities with much higher stakes in the data to use multiple passes to further prevent any possibility of it. But, we're talking reducing the possibility from near zero to even nearer to zero.

 

[Golden]

Thanks. That was very enlightening.

 

[FliGi7]

Sure thing. Sometimes it just makes more sense to kill two birds with one stone and take some aggression out by using the sledgehammer method.