process_pool

Process Pool is something I created to replace a thread pool that had major performance problems caused by Ruby's (MRI) poor thread implementation.

The basic usage is:

class SomeTask def initialize(x, y, z) end

def run end end

pool = ProcessPool.new(5) # first constructor argument is a number of worker processes pool.schedule SomeTask, 1, 2, 3 # adds task SomeTask.new(1,2,3) to the queue (instantiation happens just before calling run()) pool.start # starts the workers pool.schedule SomeTask, 3, 2, 1 # tasks can be added also after calling start pool.shutdown # waits for all the tasks in the queue to be processed and kills all the worker processes

For now there is only one, very naive, queue implementation, which uses file locking and JSON storage format. I plan on adding some other (better) queue backend soon.