Closed GoogleCodeExporter closed 8 years ago
>> On Thu, Aug 11, 2011 at 05:06:45PM -0700, Jim Garlick wrote:
>>> Hi Terence,
>>>
>>> Just returning from vacation - sorry for the delay.
>>> Can you provide a citation for the 3-pass wipe pattern?
>>>
>>> Your patch would write two passes of zeros, then verify that the
>>> disk contains the last pattern written. Is that what you want?
>>>
>>> Jim
Date: Sat, 20 Aug 2011 11:25:03 -0700
From: Terence Truong <truong.terence@gmail.com>
To: Jim Garlick <garlick@llnl.gov>
Subject: Re: Adding shorter DoD wipe pattern
Hi Jim,
I know you may still be on vacation, but I wanted to send this before I
forget.
Here is a portion of an article that I read:
DoD Short Method
When the Autonuke mode starts, the enumerated devices are automatically
sent the wipe command with the DoD Short Method. This is a triple-pass
wipe taking sequences 1, 2, and 7 from the standard DoD 5220.22-M wipe.
Active KillDisk has both the 3-pass and 7-pass option, and here is a link
to the their site:
[1]http://www.killdisk.com/commandline.htm
It is under erasemethod 2 and 3.
2 - US DoD 5220.22-M (3 passes, verify)
3 - US DoD 5220.22-M ECE (7 passes,
verify)
Thank you again.
Terence Truong
Original comment by garlick....@gmail.com
on 22 Aug 2011 at 4:00
Date: Mon, 22 Aug 2011 09:47:00 -0700
From: Jim Garlick <garlick@llnl.gov>
To: Terence Truong <truong.terence@gmail.com>
Subject: Re: Adding shorter DoD wipe pattern
Hi Terence-
OK, looking at 522.22-M (http://www.usaid.gov/policy/ads/500/d522022m.pdf)
sec 8-306, There is a 3-pass method (e) which is
1) a character
2) its complement
3) random
Is that the one you're looking for? I've tentatively added that as
"dod-short". I wonder if the killdisk documentation is referencing an
older version of 522.22-M? I don't see a 7-pass method unless I'm missing
something.
By the way, the "dod" method in scrub is my best attempt at DoD method (d)
which is supposed to be:
1) a character
2) its complement
3) random
4) verify
but I add a another fixed pattern between pass 3 and 4 because it's easier
to verify a fixed pattern than a random pattern.
Anyway, let me know what you think. I can always add the method you
asked for but I can't call it 522.22-M without a citation in taht document
for it.
Jim
Original comment by garlick....@gmail.com
on 22 Aug 2011 at 4:47
Turns out this is the same as the usarmy method.
I've documented this in the man page.
Original comment by garlick....@gmail.com
on 29 Aug 2011 at 6:02
Original issue reported on code.google.com by
garlick....@gmail.com
on 22 Aug 2011 at 3:59