Skip to content


Repository files navigation


Build Status Code Climate Gem Version Documentation Status Downloads Reviewed by Hound


Airbrake is an online tool that provides robust exception tracking in any of your Ruby applications. In doing so, it allows you to easily review errors, tie an error to an individual piece of code, and trace the cause back to recent changes. The Airbrake dashboard provides easy categorization, searching, and prioritization of exceptions so that when errors occur, your team can quickly determine the root cause.

Key features

The Airbrake Dashboard

This library is built on top of Airbrake Ruby. The difference between Airbrake and Airbrake Ruby is that the airbrake gem is just a collection of integrations with frameworks or other libraries. The airbrake-ruby gem is the core library that performs exception sending and other heavy lifting.

Normally, you just need to depend on this gem, select the integration you are interested in and follow the instructions for it. If you develop a pure frameworkless Ruby application or embed Ruby and don't need any of the listed integrations, you can depend on the airbrake-ruby gem and ignore this gem entirely.

The list of integrations that are available in this gem includes:

Deployment tracking:



Add the Airbrake gem to your Gemfile:

gem 'airbrake'


Invoke the following command from your terminal:

gem install airbrake




To integrate Airbrake with your Rails application, you need to know your project id and project key. Set AIRBRAKE_PROJECT_ID & AIRBRAKE_PROJECT_KEY environment variables with your project's values and generate the Airbrake config:


rails g airbrake

Heroku add-on users can omit specifying the key and the id. Heroku add-on's environment variables will be used (Heroku add-on docs):

rails g airbrake

This command will generate the Airbrake configuration file under config/initializers/airbrake.rb. Make sure that this file is checked into your version control system. This is enough to start Airbraking.

In order to configure the library according to your needs, open up the file and edit it. The full list of supported configuration options is available online.

To test the integration, invoke a special Rake task that we provide:

rake airbrake:test

In case of success, a test exception should appear in your dashboard.

The notify_airbrake controller helpers

The Airbrake gem defines two helper methods available inside Rails controllers: #notify_airbrake and #notify_airbrake_sync. If you want to notify Airbrake from your controllers manually, it's usually a good idea to prefer them over Airbrake.notify, because they automatically add information from the Rack environment to notices. #notify_airbrake is asynchronous, while #notify_airbrake_sync is synchronous (waits for responses from the server and returns them). The list of accepted arguments is identical to Airbrake.notify.

Additional features: user reporting, sophisticated API

The library sends all uncaught exceptions automatically, attaching the maximum possible amount information that can help you to debug errors. The Airbrake gem is capable of reporting information about the currently logged in user (id, email, username, etc.), if you use an authentication library such as Devise. The library also provides a special API for manual error reporting. The description of the API is available online.

Automatic integration with Rake tasks and Rails runner

Additionally, the Rails integration offers automatic exception reporting in any Rake tasks[link] and Rails runner.

Integration with filter_parameters

If you want to reuse Rails.application.config.filter_parameters in Airbrake you can configure your notifier the following way:

# config/initializers/airbrake.rb
Airbrake.configure do |c|
  c.blocklist_keys = Rails.application.config.filter_parameters

There are a few important details:

  1. You must load filter_parameter_logging.rb before the Airbrake config
  2. If you use Lambdas to configure filter_parameters, you need to convert them to Procs. Otherwise you will get ArgumentError
  3. If you use Procs to configure filter_parameters, the procs must return an Array of keys compatible with the Airbrake allowlist/blocklist option (String, Symbol, Regexp)

Consult the example application, which was created to show how to configure filter_parameters.

filter_parameters dot notation warning

The dot notation introduced in rails/pull/13897 for filter_parameters (e.g. a key like credit_card.code) is unsupported for performance reasons. Instead, simply specify the code key. If you have a strong opinion on this, leave a comment in the dedicated issue.


In new Rails apps, by default, all the Airbrake logs are written into log/airbrake.log. In older versions we used to write to wherever Rails.logger writes. If you wish to upgrade your app to the new behaviour, please configure your logger the following way:

c.logger = Airbrake::Rails.logger


To use Airbrake with Sinatra, simply require the gem, configure it and use our Rack middleware.

# myapp.rb
require 'sinatra/base'
require 'airbrake'

Airbrake.configure do |c|
  c.project_id = 113743
  c.project_key = 'fd04e13d806a90f96614ad8e529b2822'

  # Display debug output.
  c.logger.level = Logger::DEBUG

class MyApp < Sinatra::Base
  use Airbrake::Rack::Middleware

  get('/') { 1/0 }


To run the app, add a file called to the same directory and invoke rackup from your console.

require_relative 'myapp'

That's all! Now you can send a test request to localhost:9292 and check your project's dashboard for a new error.

curl localhost:9292


To send exceptions to Airbrake from any Rack application, simply use our Rack middleware, and configure the notifier.

require 'airbrake'
require 'airbrake/rack'

Airbrake.configure do |c|
  c.project_id = 113743
  c.project_key = 'fd04e13d806a90f96614ad8e529b2822'

use Airbrake::Rack::Middleware

Note: be aware that by default the library doesn't filter any parameters, including user passwords. To filter out passwords add a filter.

Appending information from Rack requests

If you want to append additional information from web requests (such as HTTP headers), define a special filter such as:

Airbrake.add_filter do |notice|
  next unless (request = notice.stash[:rack_request])
  notice[:params][:remoteIp] = request.env['REMOTE_IP']

The notice object carries a real Rack::Request object in its stash. Rack requests will always be accessible through the :rack_request stash key.

Optional Rack request filters

The library comes with optional predefined builders listed below.


RequestBodyFilter appends Rack request body to the notice. It accepts a length argument, which tells the filter how many bytes to read from the body.

By default, up to 4096 bytes is read:


You can redefine how many bytes to read by passing an Integer argument to the filter. For example, read up to 512 bytes:


Sending custom route breakdown performance

Arbitrary code performance instrumentation

For every route in your app Airbrake collects performance breakdown statistics. If you need to monitor a specific operation, you can capture your own breakdown:

def index
  Airbrake::Rack.capture_timing('operation name') do


That will benchmark call_operation and send performance information to Airbrake, to the corresponding route (under the 'operation name' label).

Method performance instrumentation

Alternatively, you can measure performance of a specific method:

class UsersController
  extend Airbrake::Rack::Instrumentable

  def index
  airbrake_capture_timing :index

Similarly to the previous example, performance information of the index method will be sent to Airbrake.


We support Sidekiq v2+. The configurations steps for them are identical. Simply require our integration and you're done:

require 'airbrake/sidekiq'

If you required Sidekiq before Airbrake, then you don't even have to require anything manually and it should just work out-of-box.


By default, Airbrake notifies of all errors, including reoccurring errors during a retry attempt. To filter out these errors and only get notified when Sidekiq has exhausted its retries you can add the RetryableJobsFilter:


The filter accepts an optional max_retries parameter. When set, it configures the amount of allowed job retries that won't trigger an Airbrake notification. Normally, this parameter is configured by the job itself but this setting takes the highest precedence and forces the value upon all jobs, so be careful when you use it. By default, it's not set.

Airbrake.add_filter( 10)


No additional configuration is needed. Simply ensure that you have configured your Airbrake notifier with your queue adapter.


Simply require the Resque integration:

require 'airbrake/resque'

Integrating with Rails applications

If you're working with Resque in the context of a Rails application, create a new initializer in config/initializers/resque.rb with the following content:

# config/initializers/resque.rb
require 'airbrake/resque'
Resque::Failure.backend = Resque::Failure::Airbrake

Now you're all set.

General integration

Any Ruby app using Resque can be integrated with Airbrake. If you can require the Airbrake gem after Resque, then there's no need to require airbrake/resque anymore:

require 'resque'
require 'airbrake'

Resque::Failure.backend = Resque::Failure::Airbrake

If you're unsure, just configure it similar to the Rails approach. If you use multiple backends, then continue reading the needed configuration steps in the Resque wiki (it's fairly straightforward).


Simply require our integration and you're done:

require 'airbrake/delayed_job'

If you required DelayedJob before Airbrake, then you don't even have to require anything manually and it should just work out-of-box.


Simply require our integration and you're done:

require 'airbrake/shoryuken'

If you required Shoryuken before Airbrake, then you don't even have to require anything manually and it should just work out-of-box.


Simply require our integration and you're done:

require 'airbrake/sneakers'

If you required Sneakers before Airbrake, then you don't even have to require anything manually and it should just work out-of-box.


The ActionCable integration sends errors occurring in ActionCable actions and subscribed/unsubscribed events. If you use Rails with ActionCable, there's nothing to do, it's already loaded. If you use ActionCable outside Rails, simply require it:

require 'airbrake/rails/action_cable'


Airbrake offers Rake tasks integration, which is used by our Rails integration[link]. To integrate Airbrake in any project, just require the gem in your Rakefile, if it hasn't been required and configure the notifier.

# Rakefile
require 'airbrake'

Airbrake.configure do |c|
  c.project_id = 113743
  c.project_key = 'fd04e13d806a90f96614ad8e529b2822'

task :foo do


If you want to convert your log messages to Airbrake errors, you can use our integration with Ruby's Logger class from stdlib. All you need to do is to wrap your logger in Airbrake's decorator class:

require 'airbrake/logger'

# Create a normal logger
logger =$stdout)

# Wrap it
logger =

Now you can use the logger object exactly the same way you use it. For example, calling fatal on it will both log your message and send it to the Airbrake dashboard:


The Logger class will attempt to utilize the default Airbrake notifier to deliver messages. It's possible to redefine it via #airbrake_notifier:

# Assign your own notifier.
logger.airbrake_notifier =

Airbrake severity level

In order to reduce the noise from the Logger integration it's possible to configure Airbrake severity level. For example, if you want to send only fatal messages from Logger, then configure it as follows:

# Send only fatal messages to Airbrake, ignore anything below this level.
logger.airbrake_level = Logger::FATAL

By default, airbrake_level is set to Logger::WARN, which means it sends warnings, errors and fatal error messages to Airbrake.

Configuring Airbrake logger integration with a Rails application

In order to configure a production logger with Airbrake integration, simply overwrite Rails.logger with a wrapped logger in an after_initialize callback:

# config/environments/production.rb
config.after_initialize do
  # Standard logger with Airbrake integration:
  Rails.logger =

Configuring Rails APM SQL query stats when using Rails engines

By default, the library collects Rails SQL performance stats. For standard Rails apps no extra configuration is needed. However if your app uses Rails engines, you need to take an additional step to make sure that the file and line information is present for queries being executed in the engine code.

Specifically, you need to make sure that your Rails.backtrace_cleaner has a silencer that doesn't silence engine code (will be silenced by default). For example, if your engine is called blorgh and its main directory is in the root of your project, you need to extend the default silencer provided with Rails and add the path to your engine:

# config/initializers/backtrace_silencers.rb

# Delete default silencer(s).

# Define custom silencer, which adds support for the "blorgh" engine
Rails.backtrace_cleaner.add_silencer do |line|
  app_dirs_pattern = %r{\A/?(app|config|lib|test|blorgh|\(\w*\))}

Plain Ruby scripts

Airbrake supports any type of Ruby applications including plain Ruby scripts. If you want to integrate your script with Airbrake, you don't have to use this gem. The Airbrake Ruby gem provides all the needed tooling.

Deploy tracking

By notifying Airbrake of your application deployments, all errors are resolved when a deploy occurs, so that you'll be notified again about any errors that reoccur after a deployment. Additionally, it's possible to review the errors in Airbrake that occurred before and after a deploy.

There are several ways to integrate deployment tracking with your application, that are described below.


The library supports Capistrano v2 and Capistrano v3. In order to configure deploy tracking with Capistrano simply require our integration from your Capfile:

# Capfile
require 'airbrake/capistrano'

If you use Capistrano 3, define the after :finished hook, which executes the deploy notification task (Capistrano 2 doesn't require this step).

# config/deploy.rb
namespace :deploy do
  after :finished, 'airbrake:deploy'

If you version your application, you can set the :app_version variable in config/deploy.rb, so that information will be attached to your deploy.

# config/deploy.rb
set :app_version, '1.2.3'

Rake task

A Rake task can accept several arguments shown in the table below:

Key Required Default Example
ENVIRONMENT No Rails.env production
USERNAME No nil john
REVISION No nil 38748467ea579e7ae64f7815452307c9d05e05c5
VERSION No nil v2.0

In Rails

Simply invoke rake airbrake:deploy and pass needed arguments:

rake airbrake:deploy USERNAME=john ENVIRONMENT=production REVISION=38748467 REPOSITORY=


Make sure to require the library Rake integration in your Rakefile.

# Rakefile
require 'airbrake/rake/tasks'

Then, invoke it like shown in the example for Rails.

Supported Rubies

  • CRuby >= 2.6.0
  • JRuby >= 9k


In case you have a problem, question or a bug report, feel free to:


The project uses the MIT License. See for details.

Development & testing

In order to run the test suite, first of all, clone the repo, and install dependencies with Bundler.

git clone
cd airbrake

Next, run unit tests.

bundle exec rake

In order to test integrations with frameworks and other libraries, install their dependencies with help of the following command:

bundle exec appraisal install

To run integration tests for a specific framework, use the appraisal command.

bundle exec appraisal rails-4.2 rake spec:integration:rails
bundle exec appraisal sinatra rake spec:integration:sinatra

Pro tip: GitHub Actions config has the list of all the integration tests and commands to invoke them.