System architecture¶

Design Decisions¶

In order to simplify the network protocol and scheduler, the cracker uses atomic work units. In other words, once the controller has issued a work unit to an agent, the work unit cannot be canceled or interrupted by the controller. (Once a hash is cracked, all other agents will finish their current work units.) In the event that a work unit is not completed on schedule due to a network outage, crash of the agent executable, or similar error, it will be reallocated to the next available agent.

This section is a stub. It will be expanded in the near future.

System Components¶

A complete hash cracker system consists of a single controller and one or more agents.

Controller¶

The controller coordinates the entire cracking effort, dividing search space into work units and issuing them to agents. It also is responsible for providing the web-based user interface.

Agent¶

Agents connect to the controller to receive work units, iterate through the search space, and report results.