On Mon 02 Apr 2018 at 13:07:48 (-0500), John Hasler wrote: > Curt writes: > > I guess the only means of verifying whether your data has been > > effectively destroyed is by attempting to recover it; as the > > threat-scenarios spoken about here (by individuals) generally posit > > attackers (corporate or governmental) with more resources at their > > disposal than they have at theirs... > > You also must posit the data being valuable enough to such opponents for > them to be willing to *spend* that much to get it. I'm quite confident > that none of mine is.
Agreed, but the OP is full of hypotheticals so we have to play along! > > ...this task proves to be more arduous (if not nigh on impossible) > > than the destruction itself. > > Heating the disks to well above the Curie point of the magnetic coating > is guaranteed to destroy all the data. But how to determine what the curie point of the particular drives is might be taxing. And the contents of disks have been recovered from drives in buildings consumed by fire. > The cases come apart quite easily and yield useful parts such as > rare-earth magnets. I'm quite certain that I could destroy one with a > ten-pound sledgehammer, though. Whatever the process is, it has to scale well if there are trillions of drives. So I would agree with Ben that as usual, prevention is better than cure: use LUKS from day one. Cheers, David.
