![](/i/fill.gif) |
![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Gail Shaw wrote:
>> Fact: When I was at uni, every single time assignment hand-in date came
>> around, you would not *believe* how many 3.5" disks went faulty! ;-)
>
> We had a simple solution to that problem. Faulty disk = 0%. Suddenly the
> disks got more reliable. Must have been a faulty batch.
Even more interesting: At hand-in time, there was *always* at least one
laser printer that was faulty.
[This was almost always traced back to some DILBERT putting the wrong
kind of transparency into it. Do NOT do this!!]
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
gregjohn wrote:
> "Gail Shaw" <initialsurname@sentech sa dot com> wrote:
>
>> Um, actually...
>> http://www.pcworld.com/article/id,108831-page,1/article.html
>> http://www.guard-privacy-and-online-security.com/how-clean-off-the-hard-driv
>> e.html
>>
>> Your boss is right. It may not be as easy as 'undelete' but the data's there
>> and it doesn't take a microscope to find it, and it could be worth a lot of
>> money.
>
>
> This has always bothered me. Something does not hang right.
>
> Could one just copy a couple of version of a linux distro (0.6 to 4 GB) on top
> of it? Would THAT be enough to completely obfuscate what was originally on the
> disk? If not, then why hasn't anyone yet made a commercial use of the ability
> of HDD's to remember everything that has ever been on them?
>
>
>
>
Because to go back through the information that used to be on the drive
takes advantage of the flaws in the drive it self. The read/write head
goes over the drive and writes a pattern, let's say 1111 overtop of what
used to be 0101. While the drive it sending out digital data, the
magnetic pattern on the disk doesn't exist in just a 0 or 1 state.
Think of it as just a bar refrigerator magnet, facing up/down | is one,
sideways - is 0. but the bar could be oriented like / or \. Given that
those don't appear to be past 45 degrees off vertical, I'd guess they
would both read like 1 instead of 0. Now, you have a visual of a
collection of bar magnets that aren't lined up perfectly. To write them
all to 1s, you would just rotate each one to a more vertical position.
Knowing the strength of the write head's magnet, you could guess how
much that change would be. So, you read the drive head in a less digital
and more analogue manner, and then remove the affect of the last write
phase. The more accurate the initial reading and the knowledge of the
write magnet's properties, the better you can guess past that first phase.
The drive it self just doesn't work that way. After a point, it's a
statistical guess as to how accurate the data you get back is. But some
people are willing to spend a lot of money on just a guess for certain data.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> Even more interesting: At hand-in time, there was *always* at least one
> laser printer that was faulty.
Imagine the date approaching for 300 of us to hand in our 4th year project
reports. Most reports were around 100 pages, and we needed to hand in two
copies. At the Engineering department we had 2 or 3 laser printers that
were *free* to use (as opposed to college computer rooms where you had to
pay). You can imagine the total carnage that happened the days before, I'm
sure they got in an extra pallet of A4 paper for that week.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
I can think of several ways you could "erase" the data on a hard drive:
1. Delete all the files. (E.g., delete C:\*.*)
2. Format every partition with, say, NTFS.
3. Delete all partitions.
4. Low-level format the drive (using, e.g., Maxtor PowerMax).
5. Write zeros to every byte. (E.g., dd if=/dev/zero of=/dev/sda)
6. Write random data to every byte. (...of=/dev/urandom)
7. Physically disassemble the drive, seperate the platters, magnetise
them, scratch them, smash them to pieces, and melt the pieces in a vat
of heated acid.
As I understand it, option #1 is fairly easy to reverse. (Especially if
you just delete all the files and then throw away the disk. All file
data is guaranteed to still be there, you just have to find it somehow.)
Option #2 destroys some of the filesystem metadata, but AFAIK the blocks
holding file data will still be intact. (Formatting with, say, ext2
takes a tiny fraction of the time that FAT or NTFS formatting takes, so
I'm assuming that NTFS overwrites more data. But I don't *think* it
overwrites everything.)
Option #3 should be pretty simple to reverse, AFAIK.
I am unsure as to whether #4 and #5 are different in any way. Both seem
to take the same amount of time...
As far as I know, both option #5 and option #6 make it impossible to
read any useful data from the drive just by plugging it in and asking
the drive electronics to read the data off the disk surface. In either
case, you would have to physically dismantle the drive and crawl over it
with a microscope - which is damned expensive. Using random data (and
multiple passes of it) reputedly makes this process harder - but not
impossible. Non-sequential writes are supposed to make it harder still.
As far as I can tell, option #7 gives 100% guaranteed data destruction.
But I don't have any star-drive screwdrivers. :-P [Or heated acid -
unless you count my stomach. And I am *not* putting shards of shattered
glass in there!]
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
On Fri, 27 Jun 2008 09:17:47 +0100, Invisible <voi### [at] dev null> wrote:
>I can think of several ways you could "erase" the data on a hard drive:
PGP shredder it is free
--
Regards
Stephen
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> 1. Delete all the files. (E.g., delete C:\*.*)
I wonder if you actually asked many non-geeks, what they would say if you
asked them how they would go about erasing all their data from their
computer? I guess if you try to select-all and delete the contents of drive
C from within windows it's going to get stuck at some point.
> Option #2 destroys some of the filesystem metadata, but AFAIK the blocks
> holding file data will still be intact.
Oh, I thought that "format" really went over every byte of the partition and
wrote to it? IIRC there is a "quick format" option, which I assumed just
did the headers and stuff to make it look like it was formatted.
> I am unsure as to whether #4 and #5 are different in any way. Both seem to
> take the same amount of time...
I think that writing zeros to the disk would make it easier to recover the
data, but that's just a guess.
> As far as I can tell, option #7 gives 100% guaranteed data destruction.
> But I don't have any star-drive screwdrivers. :-P [Or heated acid - unless
> you count my stomach. And I am *not* putting shards of shattered glass in
> there!]
Just chuck it on a hot fire. Saw it in half first if you want things to go
quicker.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
scott wrote:
>> 1. Delete all the files. (E.g., delete C:\*.*)
>
> I wonder if you actually asked many non-geeks, what they would say if
> you asked them how they would go about erasing all their data from their
> computer? I guess if you try to select-all and delete the contents of
> drive C from within windows it's going to get stuck at some point.
If you just want to delete "your" files, you can certainly do this. If
you want to delete the OS, you'll definitely have to do it while running
under something else. (Obviously.) I have an external USB enclosure for
this exact purpose. It's shiny. The company paid for it. :-D
>> Option #2 destroys some of the filesystem metadata, but AFAIK the
>> blocks holding file data will still be intact.
>
> Oh, I thought that "format" really went over every byte of the partition
> and wrote to it? IIRC there is a "quick format" option, which I assumed
> just did the headers and stuff to make it look like it was formatted.
Formatting a drive with ext2 just rewrites the inode tables. (Or...
something like that.) Doesn't touch most of the disk surface, and
typically takes about 20 seconds for a large partition.
A "normal" format with FAT or NTFS seems to touch more of the disk
surface, but I'm not 100% sure whether it rewrites every single block,
or just the first block in each cluster. The "quick format" option just
rewrites the file allocation table, AFAIK. (That's why it takes seconds.)
>> I am unsure as to whether #4 and #5 are different in any way. Both
>> seem to take the same amount of time...
>
> I think that writing zeros to the disk would make it easier to recover
> the data, but that's just a guess.
Writing random data is theoretically best. But in either case, once data
has been overwritten, the drive electronics can't read it back. You're
into microscope territory. The question is whether a low-level format
actually overwrites everything or not.
> Just chuck it on a hot fire. Saw it in half first if you want things to
> go quicker.
Have you ever *tried* to make a hot fire?
It's way harder than it sounds...
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Stephen wrote:
> PGP shredder it is free
1. Is it still free? IIRC they changed the licence a while back.
2. Can it erase an entire drive? I know it could do individual files,
one at a time, and I think there was a free space wipe too. But a whole
drive?
Better yet, didn't PGP used to offer a whole-drive encryption option? If
you keep your entire drive encrypted, all you need to do is delete the
encryption key. ;-)
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> Have you ever *tried* to make a hot fire?
Hot enough to melt aluminium or glass - no problem.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
On Fri, 27 Jun 2008 10:04:26 +0100, Invisible <voi### [at] dev null> wrote:
>Stephen wrote:
>
>> PGP shredder it is free
>
>1. Is it still free? IIRC they changed the licence a while back.
Shredding is free so is encrypting files and IM
>2. Can it erase an entire drive? I know it could do individual files,
>one at a time, and I think there was a free space wipe too. But a whole
>drive?
Yes it can but not with the free version, I think. I'm not going to
try it on my own HD :)
>Better yet, didn't PGP used to offer a whole-drive encryption option? If
>you keep your entire drive encrypted, all you need to do is delete the
>encryption key. ;-)
True
--
Regards
Stephen
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |