SCIENTIFIC-LINUX-USERS Archives

November 2013

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
John Pilkington <[log in to unmask]>
Reply To:
John Pilkington <[log in to unmask]>
Date:
Tue, 5 Nov 2013 10:09:43 +0000
Content-Type:
text/plain
Parts/Attachments:
text/plain (122 lines)
On 05/11/13 01:52, jdow wrote:
> On 2013/11/04 17:13, ToddAndMargo wrote:
>> On 11/04/2013 05:07 PM, Yasha Karant wrote:
>>> On 11/04/2013 04:53 PM, ToddAndMargo wrote:
>>>> On 11/04/2013 04:21 PM, Yasha Karant wrote:
>>>>> I need to do a media comparison between a data DVD and the .iso file
>>>>> that purportedly contains the image of the exact DVD (including any
>>>>> bootable or autoload binary files, not for an Intel instruction set
>>>>> architecture).
>>>>>
>>>>> When burning to the DVD, applications such as K3B and Nero (for Linux)
>>>>> will do a verify of the burned media.  My understanding is that these
>>>>> applications go through the device driver and device controller
>>>>> hardware/firmware that may be applying error correction to the raw bit
>>>>> stream; any such detected "hardware media" errors typically are
>>>>> reported
>>>>> by the driver to a log file, but typically (if corrected) do not cause
>>>>> the application to fail.
>>>>>
>>>>> If one mounts the .iso file, by a command similar to that below,
>>>>>
>>>>> # mount -t iso9660 -o ro,loop=/dev/loop0 /files/dvdimage.iso
>>>>> /media1/virtualdisc
>>>>>
>>>>> and likewise has the physical DVD in the DVD drive and mounted from,
>>>>> say, /dev/sr0
>>>>>
>>>>> will a diff /dev/loop0 /dev/sr0 suffice?
>>>>>
>>>>> Is there a utility that will do the same thing that Nero would do
>>>>> as it
>>>>> verifies after burning, but not requiring the burn -- that is,
>>>>> verify a
>>>>> DVD against an ISO image file?
>>>>>
>>>>> If /dev/sr0 were mounted on, say, /media/someDVD, and the ISO image
>>>>> file on
>>>>> /media1/virtualdisk , is there a utility or script to do a "bit by
>>>>> bit"
>>>>> comparison via the mount points (not just the "raw" mount as
>>>>> /dev/sr0 )?
>>>>>
>>>>> Yasha Karant
>>>>>
>>>>
>>>>
>>>> Hi Yasha,
>>>>
>>>> Check the DVD as a raw device.
>>>>
>>>> After you burn the ISO, eject the DVD (clears out something,
>>>> I don't know what, but had to learn the hard way):
>>>>     /usr/bin/eject /dev/sr0
>>>>
>>>> Then inject the DVD (close the door).  Can be on the same
>>>> line.
>>>>     /usr/bin/eject -t /dev/sr0
>>>>
>>>> Then make an MD5SUM of each
>>>>     md5sum /files/dvdimage.iso /dev/sr0
>>>>
>>>> Eyeball the sums.  One will be on top of the other.
>>>>
>>>> If you like, I have some leftover code I can send you.
>>>>
>>>> -T
>>>>
>>>>
>>>
>>>  From http://en.wikipedia.org/wiki/Md5sum
>>>
>>> As with all such hashing algorithms, there is theoretically an unlimited
>>> number of files that will have any given MD5 hash. However, it is very
>>> unlikely that any two non-identical files in the real world will have
>>> the same MD5 hash, unless they have been specifically created to have
>>> the same hash.
>>>
>>> End quote.
>>>
>>> I explain the above reality to my students, although I do use MD5SUM
>>> myself.  I was hoping for a utility that did a true bit-by-bit
>>> comparison of the two files.
>>>
>>>
>>> Aside:  Note that a (very) clever attacker can embed specific issues
>>> into a file such that the corrupted (and perhaps infected) file will
>>> pass a MD5 hash test.  Note that USA NSA and other entities often do
>>> employ such clever persons (do recall the cyber attack on the fissile
>>> material enrichment facilities of a Middle Eastern nation state not in
>>> full agreement with USA foreign policy, albeit an attack not
>>> specifically limited to this mechanism).  I am not suggesting that the
>>> DVD and ISO image file I am using are subject to this sort of clever
>>> corruption; but, it is important to understand the limitations of
>>> certain techniques.
>>>
>>> Yasha Karant
>>>
>>
>> Hi Yasha,
>>
>> The likely hood is pretty low.
>>
>> You could always try using the SHA sums.  Maybe do
>> both.
>>
>> -T
>
> Or use the "cmp" command, "cmp -l infile outfile" if he really wants a
> byte by byte comparison of the disks.
>
> {o.o}
>
I found this (2006) post useful several years ago when k3b verification 
was giving me trouble and there seemed to be no way of repeating just 
the verification stage.  Now I use implantisomd5 in DVDs that I create 
myself; then, with due deference to Yasha's concerns, I can reverify the 
disk later.  What to do if checkisomd5 fails is still unclear.

http://www.g-loaded.eu/2006/10/07/verify-a-burned-cddvd-image-on-linux/

John P

ATOM RSS1 RSS2