Distributed Password cracking (Restore) HELP NEEDED - URGENT!
#1
Hi Guys,

I need some urgent help, would be of great if someone can help quick. I am building a distributed system where nodes can participate for a process of cracking hash (I want to build one and not use any existing wrappers etc.). I have two cases

1. Tackling fault tolerance: Its easy as I can pass the .restore file of dead node and get it resumed from this file on new node.
2. Tackling newly joining nodes - I am stuck here. Whenever a new node joins, I don't want to keep it idle until the current cracking process is totally done. I want to assign some of the load of some current node within the network, to this new node. How can I do it?

My arguments while running for hashcat are: 
.\hashcat-3.5.0\hashcat.exe -a 3 -m 0 -s X -l Y hash ?l?l?l?l.....?l --session sessionname

My idea is to stop the current process on some running node and then assign the remaining workload to this new node and also on this running node again (Ex: A is running from -s 100 -l 1000, when the cracking process is somewhere in between, node B joins and then I want to halt A and assign the remaining workload to both A[again] and B)

With the help of restore file, can I read where exactly is the current cracking process in between -s X and -l Y and then assign the remaining load as above? an if yes, how? or else is there some other workarounds for this? I do not want to use dictionary for now and want to use this bruteforce/mask attack only.
#2
Might be advantageous to study how others have already solved this problem. The approach you described would be... less than ideal.
#3
(04-24-2017, 10:04 AM)epixoip Wrote: Might be advantageous to study how others have already solved this problem. The approach you described would be... less than ideal.

Can you help me with any references on how others solved this? Would be of great help, I am struggling a bit on this!
#4
Ill just leave this here. https://github.com/s3inlc/hashtopussy