public
Description: JobQueue means you don't have to worry about your queue no more!
Homepage: http://github.com/mloughran/job_queue/
Clone URL: git://github.com/mloughran/job_queue.git
name age message
file .gitignore Tue Feb 03 07:19:58 -0800 2009 Fix gem specification [mloughran]
file LICENSE Mon Feb 02 17:32:32 -0800 2009 Fixed the gem specifications and added README f... [mloughran]
file README.markdown Loading commit data...
file Rakefile Thu Feb 05 05:23:27 -0800 2009 Switch to using Jeweler for gem & github manage... [mloughran]
file VERSION.yml
file job_queue.gemspec
directory lib/
directory script/ Mon Feb 02 16:56:31 -0800 2009 Blank gem from newgem --simple [mloughran]
directory spec/
README.markdown

JobQueue

job_queue allows you to use lots of message queues with exactly the same interface so you don't need to worry about which queue to pick :)

This should get you started:

require 'rubygems'
require 'job_queue'

Before you can do anything you must specify an adapter to use

JobQueue.adapter = JobQueue::BeanstalkAdapter.new

Jobs can then be simply added to the queue

JobQueue.put("flubble bubble")

In your workers you'll want to subscribe to a queue

JobQueue.subscribe do |job|
  puts job
end

This subscribe block takes care of waiting for the next job to arrive and the block is passed exactly what you passed in. If you want to exit the loop just throw :stop.

JobQueue.subscribe do |job|
  # Wait - I changed my mind!
  throw :stop
end

What should you put on the queue

You might love Ruby right now, but why lock yourself in? Often the kinds of things you use queues for are the kind of things you'll want to optimize. This is a good place to start:

JSON.generate({:some => "hash"})
JSON.parse(job)

Can you show me a nice processing daemon?

Yes. Just a minute...

Adapters

Take your pick! Right now we have:

Beanstalk

http://xph.us/software/beanstalkd/

AMQP

http://github.com/tmm1/amqp/

You need to run all your code within an eventmachine loop to use AMQP.