Mailinglist Archive: opensuse (4237 mails)

< Previous Next >
re: [SLE] odd MD5sums
>> Can someone tell me if I am approaching this wrong?

>> I created .iso images of all of the original SuSE 9.1 Pro CDs and DVDs.
>> Now I am wanting to compare the MD5 checksums for the CDs with that of
>> the iso images. This worked great on the 1st 2 CDs but has gone south
>> since.

>> I am using the command:

>> md5sum /dev/cdrom

>> The odd thing is when I started getting odd readings, I ran the sum 3
>> times consecutively on the same CD and got 3 different values:
>> :~$ md5sum /dev/cdrom
>> 38e72a5efd89851ceb7349d7406ac035 /dev/cdrom
>> :~$ md5sum /dev/cdrom
>> 31dea26a1f7e5fd2c4f8a384c1bef8e0 /dev/cdrom
>> ~$ md5sum /dev/cdrom
>> 9c1e9e02ecd5401c87ac2ca59c69ee92 /dev/cdrom


>> Shouldn't I get the same value each time? Am I misunderstanding the MD5
>> sums?

Your right, something is wrong.

===> More than you wanted to know below ;-)

One thing you might try is:

dd if=/dev/cdrom | md5sum

That way dd can tell you how many blocks of data it is reading and you can confirm the whole cd is being read each time.

Finally, it you are really curious you can install dcfldd and use it to show the md5sum for each chunk of the cd.

dcfldd == DOD Computer Forensics Labs DD

I think the command would be

dcfldd if=/dev/cdrom of=/dev/null hashwindow=1M hashlog=/tmp/cdrom.hashlog

You could run that a couple of times, then diff the hashlogs to see which part of the cdrom is flackey.

one place to find dcfldd is http://sourceforge.net/project/showfiles.php?group_id=46038

Hope that was interesting if nothing else. :)

Greg
--
Greg Freemyer

< Previous Next >
Follow Ups
References