A Resque failure backend for triggering incidents in Pagerduty.

Depends on Resque 1.7 or above.

Requires a subscription to Pagerduty, with at least one Pagerduty service configured with a service type of "Generic API".

To install from Rubygems:

$ gem install resque-pagerduty

To use with bundler without adding explicit require statements to your code, add the following to your Gemfile:

gem 'resque-pagerduty', :require => 'resque_pagerduty'

Complete documentation for this gem (including this README) is available on rubydoc.info.

General information about resque failure backends is available on the resque wiki.

The Pagerduty website provides more information about the Pagerduty Integration API.

To trigger incidents in the same Pagerduty service across all jobs, configure the failure backend with a Pagerduty service key:

Resque::Failure::Pagerduty.configure do |config|
  config.service_key = 'my_pagerduty_service_key'
end

This service must be set up as a "Generic API" service in Pagerduty. The service key GUID can be found on the Pagerduty service details page.

You may want to have different jobs trigger incidents in different Pagerduty services. When handling an exception, the failure backend will automatically look for a class method named pagerduty_service_key on the resque payload class, and will preferentially use that callback. If this callback does not exist on the job class, it will default to using the service_key configured on the backend itself. For example:

class MyJob
  @queue = :my_queue

  def self.pagerduty_service_key
    'my_custom_service_key'
  end

  def self.perform(my_arg)
    # Some code that could raise an exception goes here
  end
end

If the MyJob.perform method raises an exception during processing, the failure backend would use the MyJob.pagerduty_service_key instead of the Resque::Failure::Pagerduty.service_key to trigger the incident.

When there is no Resque::Failure::Pagerduty.service_key configured, and there is no pagerduty_service_key callback defined on the job class, then the failure backend will exit gracefully. This allows you to selectively enable Pagerduty notifications for some jobs without enabling them for all jobs.

Using only the Pagerduty failure backend:

require 'resque/failure/pagerduty'

Resque::Failure::Pagerduty.configure do |config|
  config.service_key = 'my_pagerduty_service_key'
end

Resque::Failure.backend = Resque::Failure::Pagerduty

Using both the Redis and Pagerduty failure backends:

require 'resque/failure/pagerduty'
require 'resque/failure/redis'
require 'resque/failure/multiple'

Resque::Failure::Pagerduty.configure do |config|
  config.service_key = 'my_pagerduty_service_key'
end

Resque::Failure::Multiple.classes = [Resque::Failure::Redis, Resque::Failure::Pagerduty]
Resque::Failure.backend = Resque::Failure::Multiple

1. Fork the repository.
2. Create a topic branch.
3. bundle install
4. rake spec
5. Implement your feature or bug fix, including specs.
6. Add, commit, and push your changes to your fork.
7. Submit a pull request.