sentry-ruby, the Ruby Client for Sentry

The old sentry-raven client has entered maintenance mode and was moved to here.

Documentation | Bug Tracker | Forum | IRC: irc.freenode.net, #sentry

The official Ruby-language client and integration layer for the Sentry error reporting API.

Requirements

We test on Ruby 2.4, 2.5, 2.6 and 2.7 at the latest patchlevel/teeny version. We also support JRuby 9.0.

Migrate From sentry-raven

If you're using sentry-raven, we recommend you to migrate to this new SDK. You can find the benefits of migrating and how to do it in our migration guide.

Getting Started

Install

gem "sentry-ruby"

and depends on the integrations you want to have, you might also want to install these:

gem "sentry-rails"
gem "sentry-sidekiq"
# and mores to come in the future!

Sentry only runs when Sentry DSN is set

Sentry will capture and send exceptions to the Sentry server whenever its DSN is set. This makes environment-based configuration easy - if you don't want to send errors in a certain environment, just don't set the DSN in that environment!

# Set your SENTRY_DSN environment variable.
export SENTRY_DSN=http://public@example.com/project-id

# Or you can configure the client in the code.
Sentry.init do |config|
  config.dsn = 'http://public@example.com/project-id'
end

Sentry doesn't report some kinds of data by default

Sentry ignores some exceptions by default - most of these are related to 404s parameter parsing errors. For a complete list, see the IGNORE_DEFAULT constant and the integration gems' IGNORE_DEFAULT, like sentry-rails's

Sentry doesn't send personally identifiable information (pii) by default, such as request body, user ip or cookies. If you want those information to be sent, you can use the send_default_pii config option:

Sentry.init do |config|
  # other configs
  config.send_default_pii = true
end

Performance Monitoring

You can activate performance monitoring by enabling traces sampling:

Sentry.init do |config|
  # set a uniform sample rate between 0.0 and 1.0
  config.traces_sample_rate = 0.2

  # or control sampling dynamically
  config.traces_sampler = lambda do |sampling_context|
    # sampling_context[:transaction_context] contains the information about the transaction
    # sampling_context[:parent_sampled] contains the transaction's parent's sample decision
    true # return value can be a boolean or a float between 0.0 and 1.0
  end
end

To lean more about performance monitoring, please visit the official documentation.

Usage

sentry-ruby has a default integration with Rack, so you only need to use the middleware in your application like:

require 'sentry-ruby'

Sentry.init do |config|
  config.dsn = 'https://examplePublicKey@o0.ingest.sentry.io/0'

  # To activate performance monitoring, set one of these options.
  # We recommend adjusting the value in production:
  config.traces_sample_rate = 0.5
  # or
  config.traces_sampler = lambda do |context|
    true
  end
end

use Sentry::Rack::CaptureExceptions

Otherwise, Sentry you can always use the capture helpers manually

Sentry.capture_message("hello world!")

begin
  1 / 0
rescue ZeroDivisionError => exception
  Sentry.capture_exception(exception)
end

We also provide integrations with popular frameworks/libraries with the related extensions:

More configuration

You're all set - but there's a few more settings you may want to know about too!

Blocking v.s. Non-blocking

Before version 4.1.0, sentry-ruby sends every event immediately. But it can be configured to send asynchronously:

config.async = lambda { |event|
  Thread.new { Sentry.send_event(event) }
}

Using a thread to send events will be adequate for truly parallel Ruby platforms such as JRuby, though the benefit on MRI/CRuby will be limited. If the async callback raises an exception, Sentry will attempt to send synchronously.

Note that the naive example implementation has a major drawback - it can create an infinite number of threads. We recommend creating a background job, using your background job processor, that will send Sentry notifications in the background.

config.async = lambda { |event| SentryJob.perform_later(event) }

class SentryJob < ActiveJob::Base
  queue_as :default

  def perform(event)
    Sentry.send_event(event)
  end
end

After version 4.1.0, sentry-ruby sends events asynchronously by default. The functionality works like this:

When the SDK is initialized, a Sentry::BackgroundWorker will be initialized too.
When an event is passed to Client#capture_event, instead of sending it directly with Client#send_event, we'll let the worker do it.
The worker will have a number of threads. And the one of the idle threads will pick the job and call Client#send_event.

If all the threads are busy, new jobs will be put into a queue, which has a limit of 30.
If the queue size is exceeded, new events will be dropped.

However, if you still prefer to use your own async approach, that's totally fine. If you have config.async set, the worker won't initialize a thread pool and won't be used either.

About `Sentry::BackgroundWorker`

The worker is built on top of the concurrent-ruby gem's ThreadPoolExecutor, which is also used by Rails ActiveJob's async adapter. This should minimize the risk of messing up client applications with our own thread pool implementaion.

This functionality also introduces a new background_worker_threads config option. It allows you to decide how many threads should the worker hold. By default, the value will be the number of the processors your machine has. For example, if your machine has 4 processors, the value would be 4.

Of course, you can always override the value to fit your use cases, like

config.background_worker_threads = 5 # the worker will have 5 threads for sending events

You can also disable this new non-blocking behaviour by giving a 0 value:

config.background_worker_threads = 0 # all events will be sent synchronously

If you want to send a particular event immediately, you can use event hints to do it:

Sentry.capture_message("send me now!", hint: { background: false })

Contexts

In sentry-ruby, every event will inherit their contextual data from the current scope. So you can enrich the event's data by configuring the current scope like:

Sentry.configure_scope do |scope|
  scope.set_user(id: 1, email: "test@example.com")

  scope.set_tag(:tag, "foo")
  scope.set_tags(tag_1: "foo", tag_2: "bar")

  scope.set_extra(:order_number, 1234)
  scope.set_extras(order_number: 1234, tickets_count: 4)
end

Sentry.capture_exception(exception) # the event will carry all those information now

Or use top-level setters

Sentry.set_user(id: 1, email: "test@example.com")
Sentry.set_tags(tag_1: "foo", tag_2: "bar")
Sentry.set_extras(order_number: 1234, tickets_count: 4)

Or build up a temporary scope for local information:

Sentry.configure_scope do |scope|
  scope.set_tags(tag_1: "foo")
end

Sentry.with_scope do |scope|
  scope.set_tags(tag_1: "bar", tag_2: "baz")

  Sentry.capture_message("message") # this event will have 2 tags: tag_1 => "bar" and tag_2 => "baz"
end

Sentry.capture_message("another message") # this event will have 1 tag: tag_1 => "foo"

Of course, you can always assign the information on a per-event basis:

Sentry.capture_exception(exception, tags: {foo: "bar"})

More Information

Discord

dlackty/sentry-ruby