So I have a small project that I started a while back (and I mean a while) which allowed distributed brute force cracking of unix shadow passwords hashed with MD5. It is written in C and ran using MPICH 1, required a dedicated cluster and was rather not optimized. I'm planning on sitting down here soon to work on it with the following:
Now some of those features don't work well without at least a coordinated clustering effort (although working off a pre-generated table could vary from node to node, some will just be blazingly faster) but on the whole nodes would be able to come and go with central dependance on them varying by some reliablity tracker and redistribution of a range upon lack of response from a node after a period of time.
Anyways, I would like peoples input on what types of features people would like to see and if anyone is interested in helping once I get it rolling again. Cheers!
Comments
Re: Distributed Password Cracker
Excellent piece of software. Too expensive which is why we were forced to card a 20-license version (which at the time should have been enough for our needs, and was).
Secondly, this puppy needs to SUPPORT MANY MORE ENCRYPTION TYPES! PGP self-decrypting archives is an excellent start since I know of NO other software that cracks PGP. However, PGP isn't the only thing under the sun folks. More encryption types should be supported especially for those EXCRUCIATINGLY slow-to-brute algos like RAR which often run sub-100 pp/s (my god the humanity).
Otherwise very well thought out software, recommended if you can afford it (one way or another). ;) download movies
Re: Distributed Password Cracker
I have a working distributed cracker that doesn't require a dedicated cluster, is cross platform, and allows for dynamic client participation.
Check it out at http://sourceforge.net/projects/masscrack/. I would love to have some additional help optimizing it and adding hash algorithms.