A worker for Ruby background jobs
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
bin
lib
test/unit
.gems
.gems-test
LICENSE
README.md
crocodile.gemspec

README.md

Crocodile

Crocodile Gem Version

Crocodile is a gem for running periodic [background] jobs written in Ruby. Crocodile provides a worker binary that's in charge of running the jobs you define, it also provides a base class for those jobs.

Install

gem install crocodile

Jobs

Crocodile is based on some conventions for running your jobs:

  • It will look for the jobs files in the jobs directory of your application
  • The job file name is the name of the job to run
  • The job class name must be the filename ending in Job, so if your file is dummy.rb, the class defined inside the file must be DummyJob.

Your job class must respond to an interface in order to be ran by crocodile, for this, Crocodile provides the CrocodileJob class for you to inherit from there.

However, the interface is very simple, you can write your own class and forget about CrocodileJob as long as:

  • You have a class method YourClass#message that returns a string
  • You have a class method YourClass#interval that returns an integer which indicates, in seconds, how often your job will be ran
  • You have a class method YourClass#one_run_only that returns a boolean indicating if crocodile has to run it only once
  • You have a class method YourClass#logger that returns a Logger object
  • You have a class method YourClass#run that implements the logic of your job

If you use the CrocodileJob as the base class, you only need to reimplement:

  • CrocodileJob#interval
  • CrocodileJob#run

The default logger will log to STDOUT, if you want to log to a file, reimplement CrocodileJob#logger with something like Logger.new("jobs/logs/dummy.log")

Example

Suppose you have an application, let's say dummy_app and you need to run periodic background jobs on it. Go to the dummy_app dir and create the necessary directories:

mkdir -p jobs/pids

Then define a dummy job for your dummy app:

# file: jobs/dummy.rb

require 'crocodile'

class DummyJob < CrocodileJob
  def self.interval
    ENV['DUMMY_JOB_INTERVAL'] || 10
  end

  def self.run
    puts "Look ma, I'm running!!"
  end
end

Once the job is defined, go to the terminal and run it:

crocodile dummy start

You will see the output of the run method every 10 seconds. If you want to override the interval, just export the evn var:

DUMMY_JOB_INTERVAL=3 crocodile dummy start

This is useful for testing your job works correcly, however, it won't be very useful when you need to run it in a server, since you will want to log out from the server and keep the job running, so you need to daemonize your job. Also, to keep track of what your job is doing, you might want to use a log file:

# file: jobs/dummy.rb

require 'crocodile'

class DummyJob < CrocodileJob
  def self.interval
    ENV['DUMMY_JOB_INTERVAL'] || 10
  end

  def self.run
    puts "Look ma, I'm running!!"
  end

  def self.logger
    Logger.new("/tmp/dummy.log")
  end
end

Then, run it as a daemon:

crocodile -d dummy start

Now you can check the file /tmp/dummy.log for the output of your job at any given time.

The file jobs/pids/dummy.pid will show the PID of the running job.

If you need to stop the job, you do it through command line as well:

crocodile dummy stop

If you need just schedule the job instead to run it immediately, you do it through command line as well:

crocodile dummy schedule

By default job will be scheduled to start after interval.

If you want to schedule your job to start at certain hour you can redefine start_at method with format hh:mm:

class DummyJob < CrocodileJob
  def self.start_at
    "05:00"
  end
end

crocodile dummy schedule

When start_at is defined your job will start working at start_at and will execute each interval.

So, that's pretty much it. We hope you enjoy it.

How to collaborate

If you find a bug or want to collaborate with the code, you can:

  • Report issues trhough the issue tracker
  • Fork the repository into your own account and submit a Pull Request