Crash Recovery Mechanism Implementation
#1
I am not sure if this has been discussed before. However, I was thinking a feature implementation for Crash Recovery Mechanism similar to what is present in JtR would be a great addition to the oclhashcat-plus tool.

Infact, this would make it one of the best programs in the world Smile

Now, I am not sure of the difficulties to implement this mechanism. I am not sure if it is easier to implement on a CPU than on GPU. I believe since in GPU the workload gets distributed between multiple GPU cores, so saving the state would not be as easy as in a CPU.

Anyway, atom will know the best in this respect Smile
#2
You should get the point from this.

http://hashcat.net/forum/thread-1154.htm...ht=restore
#3
exactly, its no so easy to implement. as long this state remains, you can still use this workaround: https://hashcat.net/wiki/oclhashcat_plus#resume_support
#4
Thank you atom. I was late for the poll though.

yes, I understand that the implementation of resume/restore option is not easy. Infact, if there is any way to help implement that, please let me know.

The other options in the poll like Distributed Cracking don't seem easy either Smile

The workaround makes sense, however, wouldn't it limit the GPU usage? For instance, it is doing the work of a word generator and if we are piping the output to oclhashcat-plus, the GPU usage and hence the speed will come down a lot.
#5
If you amplify stuff with rules you can reach full speed
#6
Ok, I shall try it.
#7
(07-24-2012, 08:09 PM)KT819GM Wrote: http://hashcat.net/forum/thread-1154.htm...ht=restore

They vote for "Passwords > Length 15". Whats the point if you can't resume cracking 20 long passwords. Unless voters are chinese/usa government with plans to use this on a super computer.
#8
? There are few topics to vote for. One of them is Restore. Another is passwords lenght.
#9
Resume / Restore support added to latest oclHashcat-plus version. Thread closed.