A long-lived project that still receives updates
Sidekiq ultimate experience.
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
 Dependencies

Development

~> 2.0

Runtime

 Project Readme

Sidekiq::Ultimate

Sidekiq ultimate experience.


WARNING

This ia an alpha/preview software. Lots of changes will be made and eventually it will overtake sidekiq-throttled and will become truly ultimate sidekiq extension one will need. :D


Installation

Add this line to your application's Gemfile:

gem "sidekiq-ultimate"

And then execute:

$ bundle

Or install it yourself as:

$ gem install sidekiq-ultimate

Usage

Add somewhere in your app's bootstrap (e.g. config/initializers/sidekiq.rb if you are using Rails):

require "sidekiq/ultimate"
Sidekiq::Ultimate.setup!

Configuration

Resurrection event handler

An event handler can be called when a job is resurrected.

Sidekiq::Ultimate.setup! do |config|
  config.on_resurrection = ->(queue_name, jobs_count) do
    puts "Resurrected #{jobs_count} jobs from #{queue_name}"
  end
end

Resurrection counter

A resurrection counter can be enabled to count how many times a job was resurrected. If enable_resurrection_counter setting is enabled, on each resurrection event, a counter is increased. Counter value is stored in redis and has expiration time 24 hours.

For example this can be used in the ServerMiddleware later on to early return resurrected jobs based on the counter value.

enable_resurrection_counter can be either a Proc or a constant.

Having a Proc is useful if you want to enable or disable resurrection counter in run time. It will be called on each resurrection event to decide whether to increase the counter or not.

Sidekiq::Ultimate.setup! do |config|
  config.enable_resurrection_counter = -> do
    DynamicSettings.get("enable_resurrection_counter")
  end
end

Sidekiq::Ultimate.setup! do |config|
  config.enable_resurrection_counter = true
end

Read the value

Resurrection counter value can be read using Sidekiq::Ultimate::Resurrector::Count.read method.

Sidekiq::Ultimate::Resurrector::Count.read(:job_id => "2647c4fe13acc692326bd4c2")
=> 1

Empty Queues Cache Refresh Interval

Sidekiq::Ultimate.setup! do |config|
  config.empty_queues_cache_refresh_interval_sec = 42
end

Specifies how often the cache of empty queues should be refreshed. In a nutshell, this sets the maximum possible delay between when a job was pushed to previously empty queue and earliest the moment when that new job could be picked up.

Note: every sidekiq process maintains its own local cache of empty queues. Setting this interval to a low value will increase the number of Redis calls needed to check for empty queues, increasing the total load on Redis.

This setting helps manage the tradeoff between performance penalties and latency needed for reliable fetch. Under the hood, Sidekiq's default fetch occurs with a single Redis BRPOP call which is passes list of all queues to pluck work from. In contrast, reliable fetch uses LPOPRPUSH (or the equivalent LMOVE in later Redis versions) to place in progress work into a WIP queue. However, LPOPRPUSH can only check one source queue to pop from at once, and no multi-key alternative is available, so multiple Redis calls are needed to pluck work if an empty queue is checked. In order to avoid performance penalties for repeated calls to empty queues, Sidekiq Ultimate therefore maintains a list of recently know empty queues which it will avoid polling for work.

Therefore:

  • If your Sidekiq architecture has a low number of total queues, the worst case penalty for polling empty queues will be bounded, and it is reasonable to set a shorter refresh period.
  • If your Sidekiq architecture has a high number of total queues, the worst case penalty for polling empty queues is large, and it is recommended to set a longer refresh period.
  • When adjusting this setting:
    • Check that work is consumed appropriately quickly from high priority queues after they bottom out (after increasing the refresh interval)
    • Check that backlog work does not accumulate in low priority queues (after decreasing the refresh interval)

NOTICE

Throttling is brought by sidekiq-throttled and it's automatically set up by the command above - don't run Sidekiq::Throttled.setup! yourself.

Thus look up it's README for throttling configuration details.


Supported Ruby Versions

This library aims to support and is tested against the following Ruby and Redis client versions:

If something doesn't work on one of these versions, it's a bug.

This library may inadvertently work (or seem to work) on other Ruby versions, however support will only be provided for the versions listed above.

If you would like this library to support another Ruby version or implementation, you may volunteer to be a maintainer. Being a maintainer entails making sure all tests run and pass on that implementation. When something breaks on your implementation, you will be responsible for providing patches in a timely fashion. If critical issues for a particular implementation exist at the time of a major release, support for that Ruby version may be dropped.

Development

After checking out the repo, run bundle install to install dependencies. Then, run bundle exec rake spec to run the tests with ruby-rb client.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

  • Fork sidekiq-ultimate on GitHub
  • Make your changes
  • Ensure all tests pass (bundle exec rake)
  • Send a pull request
  • If we like them we'll merge them
  • If we've accepted a patch, feel free to ask for commit access!

Copyright

Copyright (c) 2018-23 SensorTower Inc.
See LICENSE.md for further details.