[ale] A little math

[Scott Castaline]

On 02/10/2010 01:08 PM, Greg Freemyer wrote:
> On Wed, Feb 10, 2010 at 12:54 PM, Brian Pitts<brian at polibyte.com>  wrote:
>> On 02/10/2010 11:35 AM, Scott Castaline wrote:
>>> Okay I've got a dumb question for you math types. How long should it
>>> take to fill a 1TB hdd with random data using
>>> "dd if=/dev/urandom of=/dev/sd"? I've done some more searching and found
>>> some numbers and if I figured them right I'm looking at about 8 days. Is
>>> this accurate?
>>
>> That sounds in the ballpark for me.
>>
>> We do a one pass wipe with random data on hard drives donated to Free IT
>> Athens. Even though most of our drives are 10 or 20 GB, on old computer
>> systems it takes hours. It doubles as a great stress test though.
>>
>
> NIST has a media sanitation guideline.
>
> For drive 20GB or larger a single pass with zeros is sufficient per
> that guideline.  They reference a NSA study on laboratory based data
> recovery showing this is sufficient.  Unfortunately I've never been
> able to get my hands on the NSA paper.
>
> My 64-bit linux boxes can wipe with zeros at about 5GB/min.
>
> Or 300 GB/hr.  So 3 hrs for a 1TB.   Maybe twice that long if your
> using a 6 or 7 yr old machine. (ie. 6 hrs).
>
> Greg
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
>
I'm just going by people recommendations in setting up a LUKS volume. 
They all agree to do the urandom data will produce a stronger means of 
encryption as opposed to all zeros, which is the last data pattern 
written in the badblocks utility.