Whenever

Whenever is a Ruby gem that provides a clear syntax for writing and deploying cron jobs.

Installation

$ gem install whenever

Or with Bundler in your Gemfile.

gem 'whenever', require: false

Getting started

$ cd /apps/my-great-project
$ bundle exec wheneverize .

This will create an initial config/schedule.rb file for you (as long as the config folder is already present in your project).

The whenever command

The whenever command will simply show you your schedule.rb file converted to cron syntax. It does not read or write your crontab file.

$ cd /apps/my-great-project
$ bundle exec whenever

To write your crontab file for your jobs, execute this command:

$ whenever --update-crontab

Other commonly used options include:

$ whenever --user app # set a user as which to install the crontab
$ whenever --load-file config/my_schedule.rb # set the schedule file
$ whenever --crontab-command 'sudo crontab' # override the crontab command

Note: If you run the whenever --update-crontab without passing the --user attribute, cron will be generated by the current user. This mean tasks that needs other user permission will fail.

You can list installed cron jobs using crontab -l.

Run whenever --help for a complete list of options for selecting the schedule to use, setting variables in the schedule, etc.

Example schedule.rb file

every 3.hours do # 1.minute 1.day 1.week 1.month 1.year is also supported
  # the following tasks are run in parallel (not in sequence)
  runner "MyModel.some_process"
  rake "my:rake:task"
  command "/usr/bin/my_great_command"
end

every 1.day, at: '4:30 am' do
  runner "MyModel.task_to_run_at_four_thirty_in_the_morning"
end

every 1.day, at: ['4:30 am', '6:00 pm'] do
  runner "Mymodel.task_to_run_in_two_times_every_day"
end

every :hour do # Many shortcuts available: :hour, :day, :month, :year, :reboot
  runner "SomeModel.ladeeda"
end

every :sunday, at: '12pm' do # Use any day of the week or :weekend, :weekday
  runner "Task.do_something_great"
end

every '0 0 27-31 * *' do
  command "echo 'you can use raw cron syntax too'"
end

# run this task only on servers with the :app role in Capistrano
# see Capistrano roles section below
every :day, at: '12:20am', roles: [:app] do
  rake "app_server:task"
end

Define your own job types

Whenever ships with three pre-defined job types: command, runner, and rake. You can define your own with job_type.

For example:

job_type :awesome, '/usr/local/bin/awesome :task :fun_level'

every 2.hours do
  awesome "party", fun_level: "extreme"
end

Would run /usr/local/bin/awesome party extreme every two hours. :task is always replaced with the first argument, and any additional :whatevers are replaced with the options passed in or by variables that have been defined with set.

The default job types that ship with Whenever are defined like so:

job_type :command, ":task :output"
job_type :rake,    "cd :path && :environment_variable=:environment bundle exec rake :task --silent :output"
job_type :runner,  "cd :path && bin/rails runner -e :environment ':task' :output"
job_type :script,  "cd :path && :environment_variable=:environment bundle exec script/:task :output"

Pre-Rails 3 apps and apps that don't use Bundler will redefine the rake and runner jobs respectively to function correctly.

If a :path is not set it will default to the directory in which whenever was executed. :environment_variable will default to 'RAILS_ENV'. :environment will default to 'production'. :output will be replaced with your output redirection settings which you can read more about here: http://github.com/javan/whenever/wiki/Output-redirection-aka-logging-your-cron-jobs

All jobs are by default run with bash -l -c 'command...'. Among other things, this allows your cron jobs to play nice with RVM by loading the entire environment instead of cron's somewhat limited environment. Read more: http://blog.scoutapp.com/articles/2010/09/07/rvm-and-cron-in-production

You can change this by setting your own :job_template.

set :job_template, "bash -l -c ':job'"

Or set the job_template to nil to have your jobs execute normally.

set :job_template, nil

Parsing dates and times

Whenever uses the Chronic gem to parse the specified dates and times.

You can set your custom Chronic configuration if the defaults don't fit you.

For example, to assume a 24 hour clock instead of the default 12 hour clock:

set :chronic_options, hours24: true

# By default this would run the job every day at 3am
every 1.day, at: '3:00' do
  runner "MyModel.nightly_archive_job"
end

You can see a list of all available options here: https://github.com/mojombo/chronic/blob/master/lib/chronic/parser.rb

Customize email recipient with the MAILTO environment variable

Output from the jobs is sent to the email address configured in the MAILTO environment variable.

There are many ways to further configure the recipient.

Example: A global configuration, overriding the environment's value:

env 'MAILTO', '[email protected]'

every 3.hours do
  command "/usr/bin/my_great_command"
end

Example: A MAILTO configured for all the jobs in an interval block:

every 3.hours, mailto: '[email protected]'  do
  command "/usr/bin/my_super_command"
end

Example: A MAILTO configured for a single job:

every 3.hours do
  command "/usr/bin/my_super_command", mailto: '[email protected]'
end

Capistrano integration

Use the built-in Capistrano recipe for easy crontab updates with deploys. For Capistrano V3, see the next section.

In your "config/deploy.rb" file:

require "whenever/capistrano"

Take a look at the recipe for options you can set. https://github.com/javan/whenever/blob/master/lib/whenever/capistrano/v2/recipes.rb For example, if you're using bundler do this:

set :whenever_command, "bundle exec whenever"
require "whenever/capistrano"

If you are using different environments (such as staging, production), then you may want to do this:

set :whenever_environment, defer { stage }
require "whenever/capistrano"

The capistrano variable :stage should be the one holding your environment name. This will make the correct :environment available in your schedule.rb.

If both your environments are on the same server you'll want to namespace them, or they'll overwrite each other when you deploy:

set :whenever_environment, defer { stage }
set :whenever_identifier, defer { "#{application}_#{stage}" }
require "whenever/capistrano"

If you use a schedule at an alternative path, you may configure it like so:

set :whenever_load_file, defer { "#{release_path}/somewhere/else/schedule.rb" }
require "whenever/capistrano"

Capistrano V3 Integration

In your "Capfile" file:

require "whenever/capistrano"

Take a look at the load:defaults task (bottom of file) for options you can set. For example, to namespace the crontab entries by application and stage do this in your "config/deploy.rb" file:

set :whenever_identifier, ->{ "#{fetch(:application)}_#{fetch(:stage)}" }

The Capistrano integration by default expects the :application variable to be set in order to scope jobs in the crontab.

Capistrano roles

The first thing to know about the new roles support is that it is entirely optional and backwards-compatible. If you don't need different jobs running on different servers in your capistrano deployment, then you can safely stop reading now and everything should just work the same way it always has.

When you define a job in your schedule.rb file, by default it will be deployed to all servers in the whenever_roles list (which defaults to [:db]).

However, if you want to restrict certain jobs to only run on subset of servers, you can add a roles: [...] argument to their definitions. Make sure to add that role to the whenever_roles list in your deploy.rb.

When you run cap deploy, jobs with a :roles list specified will only be added to the crontabs on servers with one or more of the roles in that list.

Jobs with no :roles argument will be deployed to all servers in the whenever_roles list. This is to maintain backward compatibility with previous releases of whenever.

So, for example, with the default whenever_roles of [:db], a job like this would be deployed to all servers with the :db role:

every :day, at: '12:20am' do
  rake 'foo:bar'
end

If we set whenever_roles to [:db, :app] in deploy.rb, and have the following jobs in schedule.rb:

every :day, at: '1:37pm', roles: [:app] do
  rake 'app:task' # will only be added to crontabs of :app servers
end

every :hour, roles: [:db] do
  rake 'db:task' # will only be added to crontabs of :db servers
end

every :day, at: '12:02am' do
  command "run_this_everywhere" # will be deployed to :db and :app servers
end

Here are the basic rules:

1. If a server's role isn't listed in whenever_roles, it will never have jobs added to its crontab.
2. If a server's role is listed in the whenever_roles, then it will have all jobs added to its crontab that either list that role in their :roles arg or that don't have a :roles arg.
3. If a job has a :roles arg but that role isn't in the whenever_roles list, that job will not be deployed to any server.

RVM Integration

If your production environment uses RVM (Ruby Version Manager) you will run into a gotcha that causes your cron jobs to hang. This is not directly related to Whenever, and can be tricky to debug. Your .rvmrc files must be trusted or else the cron jobs will hang waiting for the file to be trusted. A solution is to disable the prompt by adding this line to your user rvm file in ~/.rvmrc

rvm_trust_rvmrcs_flag=1

This tells rvm to trust all rvmrc files.

Heroku?

No. Heroku does not support cron, instead providing Heroku Scheduler. If you deploy to Heroku, you should use that rather than Whenever.

Testing

whenever-test is an extension to Whenever for testing a Whenever schedule.

Credit

Whenever was created for use at Inkling (http://inklingmarkets.com). Their take on it: http://blog.inklingmarkets.com/2009/02/whenever-easy-way-to-do-cron-jobs-from.html

Thanks to all the contributors who have made it even better: http://github.com/javan/whenever/contributors

Discussion / Feedback / Issues / Bugs

For general discussion and questions, please use the google group: http://groups.google.com/group/whenever-gem

If you've found a genuine bug or issue, please use the Issues section on github: http://github.com/javan/whenever/issues

Ryan Bates created a great Railscast about Whenever: http://railscasts.com/episodes/164-cron-in-ruby It's a little bit dated now, but remains a good introduction.

---

---

Copyright © 2017 Javan Makhmali