Retriable

Retriable is a simple DSL to retry failed code blocks with randomized exponential backoff time intervals. This is especially useful when interacting external APIs, remote services, or file system calls.

Table of Contents

• Requirements
• Installation
• Usage
  • Defaults
  • Options
  • Configuring Which Options to Retry With :on
  • Advanced Retry Matching With :retry_if
  • Configuration
  • Override
  • Example Usage
  • Migrating off timeout:
  • Custom Interval Array
  • Turn off Exponential Backoff
  • Callbacks
  • Ensure/Else
• Contexts
• Kernel Extension
• Testing
• Credits
• Development
  • Running Specs

Requirements

Ruby 2.3.0+

If you need ruby 2.0.0-2.2.x support, use the 3.1 branch by specifying ~3.1 in your Gemfile.

If you need ruby 1.9.3 support, use the 2.x branch by specifying ~2.1 in your Gemfile.

If you need ruby 1.8.x to 1.9.2 support, use the 1.x branch by specifying ~1.4 in your Gemfile.

Installation

Via command line:

gem install retriable

In your ruby script:

require 'retriable'

In your Gemfile:

gem 'retriable', '~> 3.8'

Usage

Code in a Retriable.retriable block will be retried if an exception is raised.

require 'retriable'

class Api
  # Use it in methods that interact with unreliable services
  def get
    Retriable.retriable do
      # code here...
    end
  end
end

Defaults

By default, Retriable will:

• rescue any exception inherited from StandardError
• make 3 tries (including the initial attempt) before raising the last exception
• use randomized exponential backoff to calculate each succeeding try interval.

The default interval table with 10 tries looks like this (in seconds, rounded to the nearest millisecond):

Retry #	Min	Average	Max
1	0.25	0.5	0.75
2	0.375	0.75	1.125
3	0.563	1.125	1.688
4	0.844	1.688	2.531
5	1.266	2.531	3.797
6	1.898	3.797	5.695
7	2.848	5.695	8.543
8	4.271	8.543	12.814
9	6.407	12.814	19.222
10	stop	stop	stop

Options

Here are the available options, in some vague order of relevance to most common use patterns:

Option	Default	Definition
tries	3	Number of attempts to make at running your code block (includes initial attempt). Pass Float::INFINITY to keep retrying until success or until max_elapsed_time is reached.
on	[StandardError]	Type of exceptions to retry. Read more.
retry_if	nil	Callable (for example a Proc or lambda) that receives the rescued exception and returns true/false to decide whether to retry. Read more.
on_retry	nil	Proc to call after each try is rescued. Pass false to disable a callback set in #configure for a single call. Read more.
sleep_disabled	false	When true, disable exponential backoff and attempt retries immediately.
base_interval	0.5	The initial interval in seconds between tries.
max_elapsed_time	900 (15 min)	The maximum amount of total time in seconds that code is allowed to keep being retried. Set to nil to disable the time limit and retry based solely on tries.
max_interval	60	The maximum interval in seconds that any individual retry can reach.
multiplier	1.5	Each successive interval grows by this factor. A multipler of 1.5 means the next interval will be 1.5x the current interval.
rand_factor	0.5	The percentage to randomize the next retry interval time. The next interval calculation is randomized_interval = retry_interval * (random value in range [1 - randomization_factor, 1 + randomization_factor])
intervals	nil	Skip generated intervals and provide your own array of intervals in seconds. Read more.
timeout	nil	Deprecated in 3.8.0 and removed in 4.0. Number of seconds to allow the code block to run before raising a Timeout::Error inside each try. nil means the code block can run forever without raising error. Non-nil values emit a deprecation warning. See Migrating off timeout:.

Timing options are validated before retrying. tries must be a positive integer when Retriable generates intervals, or Float::INFINITY for unbounded retries. base_interval, max_interval, multiplier, max_elapsed_time, and timeout must be non-negative numbers, with max_elapsed_time and timeout also accepting nil. rand_factor must be a number from 0 through 1. If provided, intervals must be an array of non-negative numbers; because it replaces generated intervals, it also overrides tries, base_interval, max_interval, rand_factor, and multiplier validation. intervals cannot be combined with tries: Float::INFINITY.

Configuring Which Options to Retry With :on

:on Can take the form:

• An Exception class (retry every exception of this type, including subclasses)
• An Array of Exception classes (retry any exception of one of these types, including subclasses)
• A Hash where the keys are Exception classes and the values are one of:
  • nil (retry every exception of the key's type, including subclasses)
  • A single Regexp pattern (retries exceptions ONLY if their message matches the pattern)
  • An array of patterns (retries exceptions ONLY if their message matches at least one of the patterns)

Advanced Retry Matching With :retry_if

Use :retry_if when retry logic depends on details that :on does not cover. The Proc receives the rescued exception and should return true to retry or false to re-raise immediately.

def caused_by?(error, klass)
  current = error
  while current
    return true if current.is_a?(klass)

    current = current.cause
  end

  false
end

Retriable.retriable(
  on: [Faraday::ConnectionFailed],
  retry_if: ->(exception) { caused_by?(exception, Errno::ECONNRESET) }
) do
  # code here...
end

:retry_if runs after the exception type has matched :on.

Configuration

You can change the global defaults with a #configure block:

Retriable.configure do |c|
  c.tries = 5
  c.max_elapsed_time = 3600 # 1 hour
end

#configure sets defaults only. Per-call options passed to Retriable.retriable and Retriable.with_context still take precedence.

Override

#with_override is a block-scoped API for forcing retry options that should take precedence over both #configure defaults and per-call options. It is primarily intended for tests — it lets a test force values like tries: 1 or base_interval: 0 so the suite runs quickly and predictably, regardless of the application's #configure defaults. In application code, prefer #configure for app-level defaults and per-call options for caller-specific values.

Retriable.with_override(tries: 1, base_interval: 0) do
  Retriable.retriable do
    # code here...
  end
end

Precedence inside the block:

with_override > local options > configure defaults

#with_override requires a block and raises ArgumentError if called without one. The override is active only while the block is executing, and is automatically restored to its previous value when the block returns or raises. Nested #with_override calls work as expected: the inner block temporarily replaces the active override and the outer override is restored when the inner block exits.

#with_override is scoped to the current thread. The active override does not affect any other thread, and child threads spawned inside the block do not inherit it. This makes #with_override safe to use in parallel test runners. Fibers running inside the same thread share the thread's active override.

#with_override stores the provided options directly. Do not mutate the options hash or nested values for the duration of the block.

For test-integration patterns (RSpec around, helper methods, Minitest, etc.), see docs/testing.md.

Example Usage

This example will only retry on a Timeout::Error, retry 3 times and sleep for a full second before each try.

Retriable.retriable(on: Timeout::Error, tries: 3, base_interval: 1) do
  # code here...
end

You can also specify multiple errors to retry on by passing an array of exceptions.

Retriable.retriable(on: [Timeout::Error, Errno::ECONNRESET]) do
  # code here...
end

You can also use a hash to specify that you only want to retry exceptions with certain messages (see the documentation above). This example will retry all ActiveRecord::RecordNotUnique exceptions, ActiveRecord::RecordInvalid exceptions where the message matches either /Parent must exist/ or /Username has already been taken/, or Mysql2::Error exceptions where the message matches /Duplicate entry/.

Retriable.retriable(on: {
  ActiveRecord::RecordNotUnique => nil,
  ActiveRecord::RecordInvalid => [/Parent must exist/, /Username has already been taken/],
  Mysql2::Error => /Duplicate entry/
}) do
  # code here...
end

Migrating off timeout:

The timeout: option is deprecated in Retriable 3.8.0 and will be removed in Retriable 4.0. It still works in 3.x, but any non-nil value supplied through Retriable.configure, Retriable.retriable(...), or Retriable.with_override(...) emits a deprecation warning. In Retriable 4.0, passing timeout: will raise ArgumentError because it will no longer be a valid option.

timeout: is deprecated because it is a thin wrapper around Timeout.timeout, which may be unsafe and even dangerous. It can interrupt the retried block at any line, including inside libraries that are not interrupt-safe.

Prefer timeout settings from the library you are calling, such as Net::HTTP#read_timeout, Net::HTTP#open_timeout, or Faraday's request timeout options. If you still need Timeout.timeout, wrap the retried block explicitly so the risk is visible at the call site:

require "timeout"

Retriable.retriable(on: Timeout::Error, tries: 3) do
  Timeout.timeout(5) do
    # code here...
  end
end

Like the deprecated timeout: option, Timeout.timeout(5) inside the block is per-try — each retry gets a fresh 5-second budget. If you want an overall cap across all retries instead, prefer max_elapsed_time:.

The deprecation warning is emitted under the :deprecated warning category and at most once per process. To silence it (for example, in tests), use the standard Ruby controls — set Warning[:deprecated] = false, run with ruby -W:no-deprecated, or override Warning.warn to filter the message.

If you need millisecond units of time for the sleep interval:

Retriable.retriable(base_interval: (200 / 1000.0)) do
  # code here...
end

Custom Interval Array

You can also bypass the built-in interval generation and provide your own array of intervals. Supplying your own intervals overrides the tries, base_interval, max_interval, rand_factor, and multiplier parameters.

Retriable.retriable(intervals: [0.5, 1.0, 2.0, 2.5]) do
  # code here...
end

This example makes 5 total attempts. If the first attempt fails, the 2nd attempt occurs 0.5 seconds later.

Unbounded Retries (Opt-in)

You can opt in to unbounded retries with tries: Float::INFINITY. This is useful for long-running worker processes where retrying should continue indefinitely, but it must be used with care.

Retriable.retriable(tries: Float::INFINITY, max_elapsed_time: 300) do
  # code here...
end

When tries: Float::INFINITY is set:

• max_elapsed_time must be a finite number. Retriable raises ArgumentError if it is nil or Float::INFINITY. This is a safety bound that prevents accidentally unbounded loops.
• Custom intervals: cannot be combined with Float::INFINITY and raises ArgumentError. Use the exponential backoff settings (base_interval, multiplier, max_interval, rand_factor) instead.

Turn off Exponential Backoff

Exponential backoff is enabled by default. If you want to simply retry code every second, 5 times maximum, you can do this:

Retriable.retriable(tries: 5, base_interval: 1.0, multiplier: 1.0, rand_factor: 0.0) do
  # code here...
end

This works by starting at a 1 second base_interval. Setting the multipler to 1.0 means each subsequent try will increase 1x, which is still 1.0 seconds, and then a rand_factor of 0.0 means that there's no randomization of that interval. (By default, it would randomize 0.5 seconds, which would mean normally the intervals would randomize between 0.5 and 1.5 seconds, but in this case rand_factor is basically being disabled.)

Another way to accomplish this would be to create an array with a fixed interval. In this example, Array.new(5, 1) creates an array with 5 elements, all with the value 1. The code block will retry up to 5 times, and wait 1 second between each attempt.

# Array.new(5, 1) # => [1, 1, 1, 1, 1]

Retriable.retriable(intervals: Array.new(5, 1)) do
  # code here...
end

If you don't want exponential backoff but you still want some randomization between intervals, this code will run every 1 seconds with a randomization factor of 0.2, which means each interval will be a random value between 0.8 and 1.2 (1 second +/- 0.2):

Retriable.retriable(base_interval: 1.0, multiplier: 1.0, rand_factor: 0.2) do
  # code here...
end

Callbacks

#retriable also provides a callback called :on_retry that will run after an exception is rescued. This callback provides the exception that was raised in the current try, the try_number, the elapsed_time for all tries so far, and the time in seconds of the next_interval. As these are specified in a Proc, unnecessary variables can be left out of the parameter list.

do_this_on_each_retry = Proc.new do |exception, try, elapsed_time, next_interval|
  log "#{exception.class}: '#{exception.message}' - #{try} tries in #{elapsed_time} seconds and #{next_interval} seconds until the next try."
end

Retriable.retriable(on_retry: do_this_on_each_retry) do
  # code here...
end

Disabling a Configured Callback Per Call

If on_retry is set in Retriable.configure, every call uses it by default. To opt a specific call out — for example, a critical call site that should not log on retry — pass on_retry: false. Passing nil does not work for this purpose because per-call options are merged over configured defaults; false is the explicit "disabled" sentinel.

Retriable.configure do |c|
  c.on_retry = ->(exception, try, elapsed_time, next_interval) { log(...) }
end

# Most calls use the configured callback.
Retriable.retriable do
  # ...
end

# This specific call opts out of the configured callback.
Retriable.retriable(on_retry: false) do
  # ...
end

Ensure/Else

What if I want to execute a code block at the end, whether or not an exception was rescued (ensure)? Or what if I want to execute a code block if no exception is raised (else)? Instead of providing more callbacks, I recommend you just wrap retriable in a begin/retry/else/ensure block:

begin
  Retriable.retriable do
    # some code
  end
rescue => e
  # run this if retriable ends up re-raising the exception
else
  # run this if retriable doesn't raise any exceptions
ensure
  # run this no matter what, exception or no exception
end

Contexts

Contexts allow you to coordinate sets of Retriable options across an application. Each context is basically an argument hash for Retriable.retriable that is stored in the Retriable.config as a simple Hash and is accessible by name. For example:

Retriable.configure do |c|
  c.contexts[:aws] = {
    tries: 3,
    base_interval: 5,
    on_retry: Proc.new { puts 'Curse you, AWS!' }
  }
  c.contexts[:mysql] = {
    tries: 10,
    multiplier: 2.5,
    on: Mysql::DeadlockException
  }
end

This will create two contexts, aws and mysql, which allow you to reuse different backoff strategies across your application without continually passing those strategy options to the retriable method.

These are used simply by calling Retriable.with_context:

# Will retry all exceptions
Retriable.with_context(:aws) do
  # aws_call
end

# Will retry Mysql::DeadlockException
Retriable.with_context(:mysql) do
  # write_to_table
end

You can even temporarily override individual options for a configured context:

Retriable.with_context(:mysql, tries: 30) do
  # write_to_table with :mysql context, except with 30 tries instead of 10
end

Kernel Extension

If you want to call Retriable.retriable without the Retriable module prefix and you don't mind extending Kernel, there is a kernel extension available for this.

In your ruby script:

require 'retriable/core_ext/kernel'

or in your Gemfile:

gem 'retriable', require: 'retriable/core_ext/kernel'

and then you can call #retriable in any context like this:

retriable do
  # code here...
end

retriable_with_context(:api) do
  # code here...
end

Testing

Retriable.with_override is designed to short-circuit retries in your test suite so failing blocks do not slow tests down. The simplest pattern is an RSpec around(:each) hook (or your test framework's equivalent) that wraps every example in with_override(tries: 1, base_interval: 0).

For Rails integration, opting out of the override for specific tests, and overriding configured contexts in tests, see docs/testing.md.

Credits

The randomized exponential backoff implementation was inspired by the one used in Google's google-http-java-client project.

Development

Running Specs

bundle exec rspec