Posts: 6
Threads: 1
Joined: Dec 2013
12-25-2013, 07:44 AM
(This post was last modified: 12-25-2013, 07:50 AM by JoeBobJr.)
I'm been testing out cuda hashcat for a while messing around cracking a wpa handshake from my router. I haven't had any issues with the cuda cracker it works great. Now I'm trying to test out just the normal hashcat that uses the cpu and for some reason it's saying
index.....: 1/328 (segments), 3050031 (words), 33550341 (bytes)
and I know for sure the file I am using has 10 million passwords in it and it looks like the cracker is saying three is only 3 mil or am I seeing the wrong thing? It's saying its going to do this full run in 2h seems weird since cuda crack to run the same file was saying 24h.
the command I am using is this maybe I am going about it wrong?
hashcat-cli64.exe -m 2500 -n 20 cap.hccap dictionaries/hash.txt
the hash.txt file is little over 10gb in size exactly 10 million passwords. I'm not sure why hashcat is saying only 3 million.
Posts: 649
Threads: 18
Joined: Nov 2010
Bet 7 million are less than 8 characters which is the minimum pass len for wpa.
Posts: 6
Threads: 1
Joined: Dec 2013
(12-25-2013, 07:49 AM)radix Wrote: Bet 7 million are less than 8 characters which is the minimum pass len for wpa.
All passwords in the file are 10 digits long. All 10 million of them.
Posts: 649
Threads: 18
Joined: Nov 2010
Hmmm... is your wordlist incremental (ie 0000000000, 0000000001, etc)? If so, try this:
./hashcat-cli64.bin -m 2500 -n 20 cap.hccap -a 3 ?d?d?d?d?d?d?d?d?d?d
Also, you have 20 processors? If not, -n should be the total number of hyperthreads, or number of cores for non hyperthreaded processors.
Posts: 6
Threads: 1
Joined: Dec 2013
12-25-2013, 08:37 AM
(This post was last modified: 12-25-2013, 08:47 AM by JoeBobJr.)
(12-25-2013, 08:19 AM)radix Wrote: Hmmm... is your wordlist incremental (ie 0000000000, 0000000001, etc)? If so, try this:
./hashcat-cli64.bin -m 2500 -n 20 cap.hccap -a 3 ?d?d?d?d?d?d?d?d?d?d
Also, you have 20 processors? If not, -n should be the total number of hyperthreads, or number of cores for non hyperthreaded processors.
No I only have 2 processors I wasn't sure what it meant by threads. I'll try it at -n 2
No it's a list starting with 8000000000 to 8999999999 just put it together real fast with crunch
Posts: 649
Threads: 18
Joined: Nov 2010
12-25-2013, 08:45 AM
(This post was last modified: 12-25-2013, 08:49 AM by radix.)
ok so you could do:
./hashcat-cli64.bin -m 2500 -n 2 cap.hccap -a 3 8?d?d?d?d?d?d?d?d?d
Saves you 10gig of space. To explain your initial paste a bit better heres the break down:
hashcat broke your wordlist into 328 segments, and is starting with the 1st, in that first segment, there are 3050031 words. In the next segments (2) there will be close to the same.
However, 3050031 words would mean only 4 segments to exhaust 10 million plains. Might be a bug, I'll look into it.
*edit*
Forgot to add a --pw-min 10 so you dont increment through all 10 positions
Posts: 6
Threads: 1
Joined: Dec 2013
(12-25-2013, 08:45 AM)radix Wrote: ok so you could do:
./hashcat-cli64.bin -m 2500 -n 2 cap.hccap -a 3 8?d?d?d?d?d?d?d?d?d
Saves you 10gig of space. To explain your initial paste a bit better heres the break down:
hashcat broke your wordlist into 328 segments, and is starting with the 1st, in that first segment, there are 3050031 words. In the next segments (2) there will be close to the same.
However, 3050031 words would mean only 4 segments to exhaust 10 million plains. Might be a bug, I'll look into it.
Oh okay that makes sense. I'll try that command you used instead of making a file and see how that works.
Posts: 6
Threads: 1
Joined: Dec 2013
By the way the main reason I am testing out the regular hashcat is I have an old mini tower it's an Acer L100 someone gave me a while back. The built in nvidia card isn't supported by cuda and I wanted to just use this machine for cracking simple stuff like this. So I don't tie up my main computer. Just trying to find a work around that would still crack at a decent rate. If the computer is just too old to crack this type of stuff I'll just give it to a relative to use for their kids or something. It uses a lot of laptop parts so I can't install a regular graphics card so I figured maybe using the cpu might work. It's dual core and has 3gb of ram it should do pretty decent at cracking.
I'd like to set it up with a file so that I can see the length of time it would take to run that 10 gb file compared to my computer running the same file on cuda to see if messing with this pc would even be worth it. I'm pretty new with this stuff and not sure how much of a difference it is between using gpu and cpu for cracking this stuff.
Posts: 6
Threads: 1
Joined: Dec 2013
(12-25-2013, 08:45 AM)radix Wrote: ok so you could do:
./hashcat-cli64.bin -m 2500 -n 2 cap.hccap -a 3 8?d?d?d?d?d?d?d?d?d
Saves you 10gig of space. To explain your initial paste a bit better heres the break down:
hashcat broke your wordlist into 328 segments, and is starting with the 1st, in that first segment, there are 3050031 words. In the next segments (2) there will be close to the same.
However, 3050031 words would mean only 4 segments to exhaust 10 million plains. Might be a bug, I'll look into it.
*edit*
Forgot to add a --pw-min 10 so you dont increment through all 10 positions
lol I noticed something wasn't right on that. It was incrementing but I thought maybe it was building a db or something to use for it. Thanks its working now.
