<img src=“https://travis-ci.org/nanodeath/threadz.png?branch=master” alt=“Build Status” /> <img src=“https://codeclimate.com/badge.png” /> <img src=“https://gemnasium.com/nanodeath/threadz.png” alt=“Dependency Status” />

This is a thread pool library that you can do two main things with, which I’ll demonstrate in code:

# These are more for "fire and forget" tasks
T1 = Threadz::ThreadPool.new
T1.process { puts "my first task" }
T1.process { puts "my second task" }

# If you care when the tasks complete, use batches
T2 = Threadz::ThreadPool.new
b = T2.new_batch
b << lambda { puts "my first task" }
b << lambda { puts "my second task" }

puts "do a couple of other things..."

b.wait_until_done

# You can do other things, too

T3 = Threadz::ThreadPool.new
b = T3.new_batch
b << lambda { puts "my first task" }
b << lambda { puts "my second task" }

puts "do a couple of other things..."

b.when_done { puts "woohoo, done with tasks" }

puts "and some other stuff, blah"

b = T3.new_batch
b << lambda { 10000000.times {} }

b.wait_until_done(:timeout => 0.1)
puts b.completed? ? "finished!" : "didn't finish"

# Error handling
b = T3.new_batch(:max_retries => 3)
b << lambda { raise }
b.wait_until_done
puts b.errors

b = T3.new_batch(:max_retries => 3, :error_handler => lambda { |error, control| puts "Error! #{error}" })
b << lambda { raise }
b.wait_until_done

# See the specs for more error handling stuff.  Much better examples.

The thread pool is also smart – depending on load, it can either spawn or cull additional threads (at a rate you can set).

For examples, please see the well-documented specs. They’re all fairly simple and straightforward. Please message me if you have issues that aren’t answered by reading the spec.