Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Mon, 29 May 2006 17:52:46 +0400
From: Solar Designer <solar@...nwall.com>
To: john-users@...ts.openwall.com
Subject: Re: is something wrong ?

On Mon, May 29, 2006 at 07:50:07AM +0200, websiteaccess@...il.com wrote:
> Benchmarking: Raw MD5 [raw-md5]... DONE
> Raw:    1109K c/s real, 1136K c/s virtual
> 
> g5-admin:/xxxx/xxxx/xxxxx/xxxxxx/JohnTheRipper/run admin$ ./john171 
> --format=raw-md5 -w:dictio.txt -rules crypt.txt
> Loaded 56 password hashes with no different salts (Raw MD5 [raw-md5])
> guesses: 0  time: 0:00:00:10 0%  c/s: 15460K  trying: isiac09
> guesses: 0  time: 0:00:00:30 1%  c/s: 15565K  trying: Gulp36
> 
>  test = 1109K c/s real, 1136K c/s virtual
>  cracking =  c/s: 15565K
> 
>  why so much diffrence ?

The c/s rate reported while cracking is the effective rate at which
combinations of {username, candidate password} are tried (even if John
happens to avoid doing those comparisons directly).  It increases along
with the number of loaded hashes per salt.  Since raw MD5 is saltless,
you can expect an effective rate of up to 60M c/s with your 56 hashes on
your CPU, but in practice this is reduced by the processing cost of
generating candidate passwords and doing the comparisons (or rather,
hash table lookups).  You're likely to get a rate somewhere closer to
60M c/s if you use "incremental" and not wordlist mode and you let it
run for a while, yet all of your hashes would remain uncracked (which is
unlikely).  If the number of hashes reduces as some are getting cracked,
the effective c/s rate will reduce too.

BTW, this is very similar to your previous question about c/s rate with
salted hashes -

	http://article.gmane.org/gmane.comp.security.openwall.john.user/537

-- 
Alexander Peslyak <solar at openwall.com>
GPG key ID: B35D3598  fp: 6429 0D7E F130 C13E C929  6447 73C3 A290 B35D 3598
http://www.openwall.com - bringing security into open computing environments

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ