Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANnLRdj4yXK-SOhzeDY3hNajokcoHNGPyijFjXqyXJq9EMaPxg@mail.gmail.com>
Date: Thu, 11 Aug 2011 12:23:16 -0600
From: Stephen John Smoogen <smooge@...il.com>
To: john-users@...ts.openwall.com
Subject: Re: JTR MPI in Virtual Machine - Stats Prob

On Thu, Aug 11, 2011 at 06:56, firstname lastname
<psykosonik_frequenz@...oo.com> wrote:
> Hello Alexander and others,

ring, if this problem is due to any of the following reasons:
>
> I am running JTR inside a Virtual Machine. It reports, 0 processors. There's no change in the real c/s value.
>
> Please let me know, where I am going wrong with this and if you need more information in order to point me in the right direction.

Well first, what are you expecting? You have a system with 1
processor. [It doesn't report 0 processors, it reports that it has one
processor in the 0 slot (C arrays start counting from 0 versus 1).] So
your aren't going to see any real increase because MPI is not going to
go out to other processors. In order for MPI or other tools to make it
go faster, they are going to need to have more processors to run on.

I believe you are seeing the virtual go up because you have created 4
processes all running at X/sec. Now I doubt they would really perform
at that long term because without additional cores/processors/etc you
are only getting any "throughput" via processor acceleration tricks
(they work short term but will shut down when the processor overheats
etc.)

> Regards,
> NeonFlash
>
>



-- 
Stephen J Smoogen.
"The core skill of innovators is error recovery, not failure avoidance."
Randy Nelson, President of Pixar University.
"Let us be kind, one to another, for most of us are fighting a hard
battle." -- Ian MacLaren

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.