EH-Net'ers:

I'm playing around with John the Ripper and the mpi patchset to use my Quad core processor to its full potential, and I have a question or two.

I've gotten the patch running using these directions:

http://hkashfi.blogspot.com/2008/12/how-to-make-johntr-use-all-of-your-cpu.html

..and all seems to be good. When I run a benchmark, I can definitely see a huge improvement in performance. However, I noticed when I use john in dictionary mode, that each processor works independently of each other. In other words, each processor cracks every password, so I get 16 results for the same entry in the password file. Does anyone know what's going on here?

Also, if I run john in incremental mode, I only get one result per entry. However, the results are not saved in a john.pot file for either mode. The best I've come up with to save my results is to just redirect standard output to a file with ">>." Does anyone have any better suggestions?

I'd really like to take a john.pot and john.rec file that was already in progress prior to my recompile with the MPI patch and pick up where it left off, but it's no big deal if that's not possible.

Thanks for any advice...

UPDATE:

Okay so, I've noticed now that the john.pot does actually get updated. It just doesn't get updated as quickly as it does on the single threaded john. It seems to be updating every 10 minutes or so. My question, though, about dictionary mode is unanswered...