The easiest way to use sidekiq without Rails

Sidekiq is usually my “goto” backend processing engine when I’m writing a Rails application.

Recently I had to write a small app that was not a web app, so it didn’t needed Rails, but I’m very comfortable with Ruby, and decided to write the background processing app, or a small “daemon” as we called it in the dark age of internet in Ruby.

And as I’m really used to the Sidekiq API and management, I used it too, but that started a small and ans easy to fix problem for me.

I wanted to use Sidekiq, and I was not using Rails, usually both are really well integrated;

After some research, I found out that you just need to use the ‘-r’ parameter from sidekiq to point it to the file it needs to require to load the workers and any other file you need it to use.

I have one project that uses only Sidekiq and ActiveRecord for example, if you think it is useful to see how to use ActiveRecord without rails, just leave a comment and I’ll write a quick tutorial.

To start our quick sample, lets create a Gemfile:

source 'https://rubygems.org'
 
gem 'sidekiq'
gem 'redis-namespace'
gem 'pry'

then we’ll create an “environment.rb” file to load the application, I’ll use an “app” directory, and make the “environment.rb” load everything that is inside the app directory, and for this sample, we’ll have an ‘workers’ directory and one ‘initializers’ to configure Sidekiq.

require 'rubygems'
require 'bundler/setup'
Bundler.require(:default)
 
$LOAD_PATH.unshift File.join(File.dirname(__FILE__),'..')
 
Dir.glob(File.join(File.dirname(__FILE__), 'app/**/*.rb')).each do |fname|
  load fname
end

with this ‘environment.rb’ file, you just need to add your code inside the app folder, and we’ll do exactly this now, creating an ‘app/initializers/sidekiq.rb’ to configure how sidekiq will connect to Redis.

Sidekiq.configure_server do |config|
  config.redis = {:namespace => 'SidekiqSample', :url => "redis:localhost:6379" }
end
 
Sidekiq.configure_client do |config|
  config.redis = {:namespace => 'SidekiqSample', :url => "redis:localhost:6379" }
end

And now we can add our workers to the app/workers directory, or any other directory we want inside app, for this sample I’ve created two workers called respectively ‘SampleWorker’ and ‘OtherWorker’, I’m posting their code bellow:

class SampleWorker
	include Sidekiq::Worker
 
	def perform(number)
		(1..number).each do |idx|
			OtherWorker.perform_async(idx)
		end
		logger.warn("Started #{number} processes")
	end
end

class OtherWorker
	include Sidekiq::Worker
 
	def perform(idx)
		logger.warn("Processing job #{idx}")
	end
end

These workers do exactly nothing, but they are good enough for this sample.

To start sidekiq, I’ve created a simple starter.sh script with the command: bundle exec sidekiq -c 2 -r ./environment.rb

And to make it easy to interact with the app, I’ve used pry to create the application console, in a ‘console’ script as you can see bellow:

#!/usr/bin/env ruby
load './environment.rb'
pry

With this, you can start your “Sidekiq only” application, and add just the libraries you need for the queue processing.

If you wanna take a look at the sample code in a more user friendly way, you can check this github repository:  https://github.com/urubatan/sidekiqsample

If you think this quick tutorial was useful, please leave a comment, and if you think this can help others, please share the link to this post to your social network.