Large IPB list issue

[M@LIK]

ok i'll try my best to describe the crazy situation : /
i got 32k hashes all iPB2[md5(md5($salt).md5($pass))]
first i ran them directly through some dicts and got positive results (~18k were found)
now i wanted to run the rest with the same dictionary BUT with rules..
and every time i try to do so i get very weird numbers::

Code:

Status.......: Running
Input.Mode...: File (___.dic)
Hash.Type....: IPB2, MyBB1.2
Time.Running.: 0 secs
Time.Left....: 1 year, 295 days
Time.Util....: 886.6ms/48.0ms Real/CPU, 5.7% idle
Speed........: 35480.1k c/s Real, 39078.4k c/s GPU
Recovered....: 0/11444 Digests, 0/10835 Salts
Progress.....: 43622785940/2026262293794090 (0.00%)
Rejected.....: 43591328660/43622785940 (99.93%)

this dict has about 154M words
notice all 99.93% rejected >.< but its decreasing

and after LONG time::

Code:

Status.......: Running
Input.Mode...: File (___.dic)
Hash.Type....: IPB2, MyBB1.2
Time.Running.: 8 hours, 50 mins
Time.Left....: 1 year, 234 days
Time.Util....: 27733711.4ms/55.3ms Real/CPU, 0.0% idle
Speed........: 39091.7k c/s Real, 39138.6k c/s GPU
Recovered....: 0/11444 Digests, 0/10835 Salts
Progress.....: 1133853248180/2026262293794090 (0.06%)
Rejected.....: 49694640820/1133853248180 (4.38%)

and the worst part is when i go with same rules and rockyou.txt::
finished after only 1:30 hours with good results @ the same speed 39Mhash/sec

and even worse when i try the run the same command again::

Code:

Status.......: Running
Input.Mode...: File (..\M\d\rockyou.dic)
Hash.Type....: IPB2, MyBB1.2
Time.Running.: 0 secs
Time.Left....: 64 days, 19 hours
Time.Util....: 544.7ms/15.0ms Real/CPU, 2.8% idle
Speed........: 33689.6k c/s Real, 39007.1k c/s GPU
Recovered....: 0/11444 Digests, 0/10835 Salts
Progress.....: 2293938450/188681672452790 (0.00%)
Rejected.....: 2275588370/2293938450 (99.20%)

>.<
i barely think everything is clear to you.. but...
any idea whats going on??

[K9]

The words could be too long after they are modified by the rules.

[M@LIK]

@K9::
then how do u explain the "Rejected" decrement along 9 hours??
and yes some rules may make it long.. but "$1" or "$1$2$3"??? : /

[undeath]

It doesn't really decrement, but there is an overflow.

The reason for having so many rejected ones appearing right after starting is simple: olchashcat-plus caches a part of the dict and at first sorts out every word to be rejected. After that it begins to run over whatever is left of that cached chunk.

[atom]

yes that explains the rejects, but it still fullfills the buffer till its full or EOF before launching the kernel

[M@LIK]

@atom::
sorry pal.. but i dont understand.. could u explain a little bit more : )
and the thing i really want to have explained is how i was able to run rockyou which has 14M words without any rejection and it finished in 1 hour! with the same rules.. problem is that i didnt keep any kind of log of printscreens to determine the difference >.<

[atom]

While it buffers the words from the dictionary it checks if a word is undersized or oversized. If this is the case it adds for each word the number of RULES multiplied by the number of SALTS to the rejection counter.

[M@LIK]

ok.. whats the solution??
split the hash file, sort the dict by length, remove the rules??

i already tried sorting the dict by length.. and it started with ZERO rejected but the same long numbers : /

[atom]

yeah, split the hash file. in theory you can split by 1 since it will not make any speed difference in sum. multihash support for salted hashes has just been added because of the lazyness of some users

[M@LIK]

@atom:: hmmm... ok lets say i can split the file by 1.. and add them to a batch file somehow,, don't think it won't so efficient due to reloading the dict and all with every single hash??
appreciate the help : )