Hashtopolis Forum

Full Version: client doesn´t "start" chunk
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Current Server Version
0.10.1
Current Client Version
0.4.0.1
Hashcat Version
5.0.1

Task command you are using
"#HL# -a 3  -1 äöüß -2 ?l?1 -3 ÄÖÜ -4 ?u?3 ?d -O "


(Thats a subtask of a supertask. I use a mask like this one:

äöüß,?l?1,ÄÖÜ,?u?3,?d

äöüß,?l?1,ÄÖÜ,?u?3,?d?d
äöüß,?l?1,ÄÖÜ,?u?3,?2
äöüß,?l?1,ÄÖÜ,?u?3,?d?d?d?d
äöüß,?l?1,ÄÖÜ,?u?3,?d?d?d?d?d?d
äöüß,?l?1,ÄÖÜ,?u?3,?d?d?d?d?d
äöüß,?l?1,ÄÖÜ,?u?3,?2?2
äöüß,?l?1,ÄÖÜ,?u?3,?d?d?d)

and so on)

Currently i have the server running in a VBox VM under lubuntu (LTS Version).
I have 5 Clients running. With two of them i have the following problem:

when i activate only a single client (with this specific problem) everything works fine.
After i activated two other clients in addition, the first one skips "starting chunk". 
It simply loops:

got Task with ID XY
Dowloading [=====]
Client is up to date
got cracker binary type hashcat
Benchmark test...
Server accepted benchmark!
got Task with ID XY
Dowloading [=====]
Client is up to date
got cracker binary type hashcat

and so on...

after i deactivate the additional clients, he starts chunks again.

the debug output is in the attachment.

thanks for your help!
The problem with your masks is that they are quite small, therefore one agent easily can complete the task in short time (I think in that case even in a matter of seconds). Because each agent is benchmarking the task and then tries to get a chunk, only one agent "wins" and can complete the task. Of course this is a large overhead and completely useless.

In case you are importing the tasks via the mask import (for supertasks), select the 'Is Small' flag. If you create the tasks manually, you can select this on the task creation. This sets, that only one agents completes the corresponding tasks and the other agents automatically move on. This avoids multiple agents trying to work on one task and additionally, there will be no benchmarking of the task (as the agent simply will get the full task).
This worked for me.
I didn´t check this box because further down in the list, bigger tasks appear.
I think i´ll have to split the supertask then into 2 pieces!?

Thank you so far.
I think that might be the best approach then to split it into two supertasks.