Cacher is the code snippet organizer for pro developers

What is Resque?

• Handles background jobs by placing them on queues and handling them at a later time
• Useful for code that is being executed on an external web service that can be moved into a background task
  • Why?
    • The service could take up time and cause delay in your app's loading time
    • You want to move them into a background process
    • Also prevents the blocking of other requests from one that is currently being held

Setting up Resque

• install or have installed redis!!!
• start the redis process!
• In Gemfile: add gem 'resque', then run bundle install
• Add a file within lib/tasks, called resque.rake

require 'resque/tasks'

#load the environment when the worker starts up
task "resque:setup" => :environment

• run this in the terminal after all changes have been made to the project: bundle exec rake resque:work QUEUE='*'
• The QUEUE='' takes in the name of the queue to work on or use * to load all queues

---

Steps

• To add a job to a queue:

  • Use Resque.enqueue(workerClassName, additional_arguments) in your code
  • Additional arguments could be @object.id (can be use to refetch the object without submitting a complex object that is going to be turned into JSON).

• Create a new directory within apps called workers, for ex: /lib/workers

• Create a file (worker) in the workers directory and name it using the worker_class_name.rb used in the enqueue line above.

  • That worker file is basically a ruby class!

  class WorkerClassName
    # 1st requirement
    @queue = :name_of_queue       #can be named anything
    
    # 2nd requirement
    def self.perform(object.id)      #takes the additional arguments passed in to enqueue
      object = Object.find(abject.id) #refetch the id that was before to complex to send 
                                      #through
      ... #code referencing your API call!
    end
    
  end
  

• Complete!

• If your worker needs to reference a rake task: Rake::Task['file:name_of_task'].invoke

---

Resque Web Interface

• from terminal, execute resque-web
• You can check the queues and failed jobs!
• May need to restart rake task if new code was added after running the task initially

Imbedding the Web Interface into your App?

• You can add a route to also visit the resque web interface instead of launching it from terminal!
• In config/routes: mount Resque::Server, :at => "/resque"
• Need to edit the Gemfile again: gem 'resque', :require => "resque/server"

Add authentication to the Resque Web Interface

• If using Devise:

#routes.rb
authenticate :admin do
  mount Resque::Server, :at => "/resque"
end

• If not using Devise, an alternative way is to create a new initializer in config/initializers called resque_auth.rb

Resque::Server.use(Rack:;Auth::Basic) do |user, password|
  user == "rtrw"
  password == "jhtrw"
end

• A resque job is any Ruby class or module with a perform class method.

• Resque can maintain multiple queues for different job types: by setting the @queue instance var, this worker will only look for jobs on the :sleep queue!

• Next, queue the job: Resque.enque(Sleeper, 5) (in your controller, for example)

  • The parameters will be serialized as JSON and appended to the Redis queue specified in the job class:

  {
    'class': 'Sleeper',
    'args': [5]
  }
  

• Jobs should only need access to your models!

• When a job is created, it gets appended to a list data structure in Redis.

  • A resque worker will then try to process the job.

---

Resque Web Backend

Overview

• Contains list of all the queues and workers.
• Each queue shows a count of pending jobs.
• The list of workers displays what queue the worker is working on and the job currently being processed on.

Failed

• Shows a list of errors of the failed jobs

---

Starting up the Workers

• At the top of the your Rakefile, add the following: require 'resque/rake'
  • Run bundle exec rake -T resque
• Environment Variables:
  • QUEUE: controls WHICH queue is monitored
  • COUNT: sets the number of workers (only resque:workers)
• Start the task by running: bundle exec rake environment resque:work QUEUE=sleep
  • resque:work starts a rescue worker (more info run the rake command above)