Re: Default number of overwrites in shred

[Peter Eckersley]

I don't think that any of the attacks Wietse discussed are related to
actually reading things out from underneath multiple disk-write cycles
(though he does cite Peter Gutmann's 1996 paper on microscopic attacks,
but the paper applied to older hard disks).

Everything else is about keeping track of the behaviour of the
filesystem code, and making sure the writes hit the disk in the right
place [*].  I haven't audited the shred code, but I hope Wietse's bugs
have been fixed!

[*] One nasty gotcha there is the issue of the disk firmware moving
blocks around without telling the operating system at all.  I guess in
an ideal world, chattr +s would get the OS to tell the firmware to
always shred the blocks in that file.


On Thu, 2007-05-03 at 23:54 +0100, James Youngman wrote:
> On 5/3/07, Peter Eckersley <address@hidden> wrote:
> 
> > We're certainly working on making sure that userspace chattr()s the
> > attribute.  Apparently, for many journalling filesystems (ext3 and maybe
> > Reiser, IIRC), shred should still work pretty well as-is -- I guess
> > that's because the filesystem promises metadata consistency, not data
> > consistency.  For the others, we can certainly ping the developers about
> > it.  We could potentially work on patches too, but that's less likely
> > unless there's an obvious case where the patch would make a lot of
> > real-world difference.
> 
> There is a very interesting presentation by Wietse Venema which
> discusses this in detail (and, sadly, demonstrates that it is much
> harder to do this reliably that one might think).
> I can't find a link to the talk but there is an HTML version of it in
> the Google Cache at
> 
> http://66.102.9.104/search?q=cache:fNI_tJkUHzIJ:www.hack.lu/images/8/80/Venema.ppt+Venema.ppt&hl=en&ct=clnk&cd=3
> 
> I think his comments about this relate mainly to ZFS and ext3, though.
> 
> James.
-- 
Peter Eckersley                            address@hidden
Staff Technologist                Tel  +1 415 436 9333 x131
Electronic Frontier Foundation    Fax  +1 415 436 9993