Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 13 May 2012 20:46:38 +0530
From: Dhiru Kholia <dhiru.kholia@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: SSH thread-safety

On Sun, May 13, 2012 at 8:02 PM, Solar Designer <solar@...nwall.com> wrote:
> On Sun, May 13, 2012 at 02:41:01PM +0530, Dhiru Kholia wrote:
>> One more observation, during actual cracking the speed is quite low
>> compared to benchmarks,
>>
>> $ ../run/john -i sample
>> Loaded 1 password hash (ssh [32/64])
>> guesses: 0  time: 0:00:04:01 0.00%  c/s: 16633  trying: hbna
>
> Hmm, that's a lot lower than we've seen in actual cracking before.

This question is now resolved. The sample key I posted numbers for has
higher key length (2048) than the one in self-test. So the above
posted speeds are "normal".

>> Any ideas what is going on?
>
> Maybe some change in OpenSSL?  Our previous benchmarks on bull were on
> March 18, but OpenSSL has since been upgraded.

My machine has latest OpenSSL (1.0.1b 26 Apr 2012) while bull has
version 1.0.1 14 Mar 2012. I think that the OpenSSL version doesn't
explain the difference in speed. This needs further investigation.

> What if you add more test vectors - will the benchmark show poor speeds
> like above?

I have added a self-test key with key length of 2048 bits (since the
man page says "For RSA keys, the minimum size is 768 bits and the
default is 2048 bits.") . Doing this has slowed down the benchmark
considerably.

$ ../run/john -format=ssh -t
Benchmarking: ssh [32/64]... DONE
Raw:	19988 c/s real, 19988 c/s virtual

-- 
Cheers,
Dhiru

Powered by blists - more mailing lists

Your e-mail address:

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