Sidekiq performs background jobs. While its threading model allows it to scale easier than worker-pre-process background systems, people running test or lightly loaded systems on Heroku still want to scale down to zero to avoid racking up charges.

Tested on Ruby 1.9.2 and Heroku Cedar stack.

gem install autoscaler

This gem uses the Heroku-Api gem, which requires an API key from Heroku. It will also need the heroku app name. By default, these are specified through environment variables. You can also pass them to HerokuScaler explicitly.

HEROKU_API_KEY=.....
HEROKU_APP=....

This gem assumes the name of the Heroku process to be 'worker' when it invokes the logic to scale Sidekiq up and down. You can optionally customize the name of the process with the following environment variable

SIDEKIQ_WORKER_NAME=....

Install the middleware in your Sidekiq.configure_ blocks

require 'autoscaler/sidekiq'
require 'autoscaler/heroku_scaler'

Sidekiq.configure_client do |config|
  config.client_middleware do |chain|
    chain.add Autoscaler::Sidekiq::Client, 'default' => Autoscaler::HerokuScaler.new
  end
end

Sidekiq.configure_server do |config|
  config.server_middleware do |chain|
    chain.add(Autoscaler::Sidekiq::Server, Autoscaler::HerokuScaler.new, 60) # 60 second timeout
  end
end

• HerokuScaler includes an attempt at current-worker cache that may be overcomplication, and doesn't work very well on the server
• Multiple scale-down loops may be started, particularly if there are multiple jobs queued when the servers comes up. Heroku seems to handle multiple scale-down commands well.
• The scale-down monitor is triggered on job completion (and server middleware is only run around jobs), so if the server nevers processes any jobs, it won't turn off.
• The retry and schedule lists are considered - if you schedule a long-running task, the process will not scale-down.
• If background jobs trigger jobs in other scaled processes, please note you'll need config.client_middleware in your Sidekiq.configure_server block in order to scale-up.
• Exceptions while calling the Heroku API are caught and printed by default. See HerokuScaler#exception_handler to override

You can pass a scaling strategy object instead of the timeout to the server middleware. The object (or lambda) should respond to #call(system, idle_time) and return the desired number of workers. See lib/autoscaler/binary_scaling_strategy.rb for an example.

Client#set_initial_workers to start workers on main process startup; typically:

Autoscaler::Sidekiq::Client.add_to_chain(chain, 'default' => heroku).set_initial_workers

scaler.counter_cache = Autoscaler::CounterCacheRedis(Sidekiq.method(:redis))

The project is setup to run RSpec with Guard. It expects a redis instance on a custom port, which is started by the Guardfile.

The HerokuScaler is not tested by default because it makes live API requests. Specify HEROKU_APP and HEROKU_API_KEY on the command line, and then watch your app's logs.

HEROKU_APP=... HEROKU_API_KEY=... guard
heroku logs --app ...

Justin Love, @wondible, https://github.com/JustinLove

Ported to Heroku-Api by Fix Peña, https://github.com/fixr

Retry/schedule sets by Matt Anderson https://github.com/tonkapark and Thibaud Guillaume-Gentil https://github.com/jilion

Released under the MIT license.