identical disk dump

[John Summerfied]

Andrew Schaffner wrote:
> I have an admin question:
> 
> I need to make an identical bit-for-bit transfer of a hard disk from a
> *REALLY* slow box to a faster box.
> 
> 1- There are no security issues --plugging one box directly into the other.
> 2- I need to prove/document the copy is identical to the original.

Good grief! how does one document that _any_ program does what its 
documentation says? Especially, considering that a program and its 
document often differ.

What if the kernel does something bad? Just for instance, I cam imagine 
the process of copying a file and writing an encrypted version of it. 
Someone trying to compare the files would likely conclude they were 
different as, technically, they are.

At some point, someone has to either test the results for themselves, or 
accept your word that you employed some particular process.

_I_ am happy to believe dd does what users, having read its manual, 
would reasonably expect it does for those case in which I have use it. 
My belief is informed by my regular use of it over some years, and by 
the lack of reports of its being unreliable.

> 3- I want to do it as efficiently as possible.
> 4- I "own" both boxes.
> ---
> Has anyone used dcfldd ?? Is this a better tool than dd ??
> http://dcfldd.sourceforge.net/

I myself have never used it, though I can imagine I might in order to 
duplicate some disk drives, but then duplicating drives involves a 
certain amount of fiddling around and a mass roll-out in my environment 
would be best done by installing the drives in the recipient computers 
and booting a Linux CD to copy data off a server, or in some cases off 
the CD.


> ---
> Any time-tested commands/flags/scripts would be greatly appreciated.
> 
> Best wishes to everyone,
> 
> Andrew
> 
> 


-- 

Cheers
John

-- spambait
1aaaaaaa at computerdatasafe.com.au  Z1aaaaaaa at computerdatasafe.com.au
Tourist pics http://portgeographe.environmentaldisasters.cds.merseine.nu/

do not reply off-list