Tag: concurrency

Ruby concurrency is hard: how I became a Ruby on Rails contributor

For the past several weeks, I've been trying to fix a cranky spec in Karafka integrations suite, which in the end, lead me to become a Ruby on Rails micro-contributor and submitting similar fix to several other high-popularity projects from the Ruby ecosystem. Here's my story of trying to make sense of my specs and Ruby concurrency.

Ephemeral bug from a test-suite

Karafka is a Ruby and Rails multi-threaded efficient Kafka processing framework. To provide reliable OSS that is multi-threaded, I had to have the option to run my test suite concurrently to simulate how Karafka operates. Since it was a specific use case, I created my micro-framework.

Long story short: It runs end-to-end integration specs by running them in separate Ruby processes. Each starts Karafka, runs all the code in various configurations, connects to Kafka, checks assertions, and at the end, shuts down.

Such an approach allowed me to ensure that the process's whole lifecycle and its components work as expected. Specs are started with supervision, so in case of any hang, it will be killed after 5 minutes.

Karafka itself also has an internal shutdown supervisor. In case of a user shutdown request, if the shutdown takes longer than the defined expected time, Karafka will stop despite having things running. And this is what was happening with this single spec:

E, [2022-11-19T16:47:49.602718 #14843] ERROR -- : Forceful Karafka server stop
F, [2022-11-19T16:47:49.602825 #14843] FATAL -- : #<Karafka::Core::Monitoring::Event:0x0000562932d752b0 @id="error.occurred", @payload={:caller=>Karafka::Server, :error=>#<Karafka::Errors::ForcefulShutdownError: Karafka::Errors::ForcefulShutdownError>, :type=>"app.stopping.error"}>

This damn spec did not want to stop!

Many things are working under the hood:

  • workers that process jobs that could hang and force the process to wait
  • jobs queue that is also connected to the polling thread (to poll more data when no work is to be done)
  • listeners that poll data from Kafka that could hang
  • consumer groups with several threads polling Kafka data that might get stuck because of some underlying error
  • Other bugs in the coordination of work and states.

One thing that certainly worked was the process supervision that would forcefully kill it after 30 seconds.

Process shutdown coordination

The graceful shutdown of such a process takes work. When you have many connections to Kafka, upon a poorly organized shutdown, you may trigger several rebalances that may cause short-lived topics assignments causing nothing except friction and potentially blocking the whole process.

To mitigate this, Karafka shuts down actively and gracefully. That is, until the absolute end, it claims the ownership of given topics and partitions, actively waiting for all the current work to be finished. This looks more or less like so:

Note: Consumer groups internally in Karafka are a bit different than Kafka consumer groups. Here we focus on internal Karafka concepts.

Pinpointing the issue

After several failed attempts and fixing other bugs, I added a lot of extra instrumentation to check what Karafka hangs on. It was hanging because there were hanging listener threads!

As stated above, to close Karafka gracefully, all work from the jobs queue needs to be finished, and listeners that poll data from Kafka need to be able to exit the polling loops. It's all coordinated using a job queue. The job queue we're using is pretty complex with some blocking capabilities, and you can read about it here, but the interesting part of the code can be reduced to this:

@semaphores = Concurrent::Map.new { |h, k| h[k] = Queue.new }

Those queues are used as semaphores in the polling loops until all the current work is done. Since each Queue is assigned to a different subscription group within its thread and hidden behind a concurrent map, there should be no problem. Right?


Once I had my crazy suspicion, I decided to reduce it down to a proof of concept:

require 'concurrent-ruby'

100.times do
  ids = Set.new
  semaphores = Concurrent::Hash.new { |h, k| h[k] = Queue.new }

  100.times.map do
    Thread.new do
      ids << semaphores['test'].object_id

  raise "I expected 1 semaphore but got #{ids.size}" if ids.size != 1

once executed, boom:

poc.rb:13:in `<main>': I expected 1 semaphore but got 2 (RuntimeError)

There is more than one semaphore for one listener! This caused Karafka to wait until forced to stop because the worker thread would use a different semaphore than the listener thread.

But how is that even possible?

Well, Concurrent::Hash and Concurrent::Map initialization is indeed thread-safe but not precisely as you would expect them to be. The docs state that:

This version locks against the object itself for every method call, ensuring only one thread can be reading or writing at a time. This includes iteration methods like #each, which takes the lock repeatedly when reading an item.

"only one thread can be reading or writing at a time". However, we are doing both at different times. Our code:

semaphores = Concurrent::Hash.new { |h, k| h[k] = Queue.new }

is actually equivalent to:

semaphores = Concurrent::Hash.new do |h, k|
  queue = Queue.new
  h[k] = queue

and the block content is not locked fully. One threads queue can overwrite the other if the Ruby scheduler stops the execution in the middle. Here's the flow of things happening in the form of a diagram:

Once in a while listener would receive a dangling queue object, effectively blocking the polling process.

Fixing the issue

This can be fixed either by replacing the Concurrent::Hash with Concurrent::Map and using the #compute_if_absent method or by introducing a lock inside of the Concurrent::Hash initialization block:

Concurrent::Map.new do |k, v|
  k.compute_if_absent(v) { [] }

mutex = Mutex.new

Concurrent::Hash.new do |k, v|
  mutex.synchronize do
    break k[v] if k.key?(v)
    k[v] = []

Okay, but what does Ruby on Rails and other projects do with all of this?

Fixing the world

I figured out that if I made this mistake, maybe others did. I decided to check my local gems to find occurrences quickly. Inside my local gem cache, I executed the following code:

fgrep -R 'Concurrent::Hash.new {' ./
fgrep -R 'Concurrent::Hash.new do' ./
fgrep -R 'Concurrent::Map.new {' ./
fgrep -R 'Concurrent::Map.new do' ./

and validated that I'm not an isolated case. I wasn't alone!

Then using Sourcegraph I pinpointed a few projects that had the potential for fixes:

  • rails (activesupport and actionview)
  • i18n
  • dry-schema
  • finite_machine
  • graphql-ruby
  • rom-factory
  • apache whimsy
  • krane
  • puppet

I am not a domain expert in any of those, and understanding the severity of each was beyond my time constraints, but I decided to give it a shot.

Rails (ActiveSupport and ActionView)

Within Rails, this "pattern" was used twice: in ActiveSupport and ActionView.

In ActionView, it was used within a cache:

PREFIXED_PARTIAL_NAMES = Concurrent::Map.new do |h, k|
  h[k] = Concurrent::Map.new

and assuming that the cached result is stateless (same result each time for the same key), the issue could only cause an extra computation upon first parallel requests to this cache.

In the case of ActiveSupport, none of the concurrency code was needed, so I just replaced it with a simple Hash.

Both, luckily, were not that severe, though worth fixing nonetheless.

PR: https://github.com/rails/rails/pull/46536
PR: https://github.com/rails/rails/pull/46534

Both were merged, and this is how I became a Ruby on Rails contributor :)


This case was slightly more interesting because the concurrent cache stores all translations. In theory, this could cause similar leakage as in Karafka, effectively losing a language by loading it to a different Concurrent::Hash:

100.times.map do
  Thread.new do
    I18n.backend.store_translations(rand.to_s, :foo => { :bar => 'bar', :baz => 'baz' })

I18n.available_locales.count #=> 1

This could lead to hard-to-debug problems. Once in a while, your system could raise something like this:

:en is not a valid locale (I18n::InvalidLocale)

without an apparent reason, and this problem would go away after a restart.

PR: https://github.com/ruby-i18n/i18n/pull/644


Another cache case where the risk would revolve around double-computing.

PR: https://github.com/dry-rb/dry-schema/pull/440


This one is interesting! Let's reduce the code to a smaller POC first and see what will happen under heavy threading:

require 'singleton'
require 'concurrent-ruby'

class Sequences
  include Singleton

  attr_reader :registry

  def initialize

  def next(key)
    registry[key] += 1

  def reset
    @registry = Concurrent::Map.new { |h, k| h[k] = 0 }

seq = Sequences.instance

loop do
  100.times.map do
    Thread.new { seq.next('boom') }

  size = seq.registry['boom']

  raise "Wanted 100 but got #{size}" unless size == 100

poc.rb:37:in `block in <main>': Wanted 100 but got 1 (RuntimeError)

The counter value gets biased. What is even more interesting is that making the map safe won't be enough:

@registry = Concurrent::Map.new { |h, k| h.compute_if_absent(k) { 0 } }
poc.rb:36:in `block in <main>': Wanted 100 but got 55 (RuntimeError)

there is one more "unsafe" method:

def next(key)
  registry[key] += 1

this operation also is not atomic, thus needs to be wrapped with a mutex:

def initialize
  @mutex = Mutex.new

def next(key)
  @mutex.synchronize do
    registry[key] += 1

Only then is this code safe to be used.


Other repositories


In my opinion, there are a few outcomes of this story:

  • Karafka has a solid test-suite!
  • If you are doing concurrency-related work, you better test it in a multi-threaded environment and test it well.
  • Concurrency is hard to many of us (maybe that's because we are special ;) ).
  • RTFM and read it well :)
  • Do not be afraid to help others by submitting pull requests!

On the other hand, looking at the frequency of this issue, it may be worth opening a discussion about changing this behavior and making the initialization fully locked.


Concurrent::Hash under cRuby is just a Hash. You can check it out here.

Cover photo by James Broad on Attribution-NonCommercial-ShareAlike 2.0 Generic (CC BY-NC-SA 2.0). Image has been cropped.

Sidekiq Unique Jobs: don’t waste your time waiting – reschedule if busy

Half a year ago, I've made a post about Enforcing unique jobs in Karafka and Sidekiq for single resources. This approach is great, however, there's a particular case in which Sidekiq Unique Jobs can block all of your Sidekiq workers except one. This can significantly limit your computing power without you being aware of it.

When using Sidekiq Unique Jobs with a WhileExecuting strategy, only a single worker can start processing Sidekiq job with a given unique key. This is really helpful when you work with resources for which you cannot perform parallel operations (for example when you work with a database for which there are no atomic operations but you need to increment counters), as doing so could overwrite results from an other worker. A WhileExecuting strategy, with a properly defined unique key can help you prevent that from happening. However...

Problem definition

Sidekiq jobs are being consumed out of a FIFO queue. Without any additional modifications, situation is pretty clear: having 4 single threaded processes each allows you to process 4 jobs at the same time.

Everything changes, when you decide to add Sidekiq Unique Jobs to your stack. In a case, when there are multiple jobs in sequence for a given unique key, Sidekiq will get seriously clogged. In the worst scenario, it won't matter how many workers and threads you have in your Sidekiq infrastructure. You will get a performance of a single Sidekiq thread. It is because you cannot make Sidekiq skip certain tasks because of FIFO queue processing. To bypass this limitation, authors of Sidekiq Unique Jobs introduced a #sleep that will run up until the resource is free again to be processed or until timeout occurs. This approach means, that if you have more tasks in queue than processors, they will have to wait until all the jobs with a given unique key are processed.

All the workers will actively wait (meaning that in Sidekiq console you will see them marked as busy) up until a lock is released.

Solution: reschedule instead of waiting

Warning: if a similar case occurs in your business logic quite often, probably you will be better taking engine different than Sidekiq. I would recommend this solution for non-frequent edge/corner cases.

Bypassing that behavior is pretty easy: if there's a lock, put the current job at the end of the queue. That said, your jobs will be checked for possibility of execution and rescheduled back instead of waiting. This will mess up your queue counters a bit (as you will have more jobs enqueued and processed that it should) but on the other hand it means that Sidekiq will "actively" seek for resources on which it can work in a certain moment.

To do so, we can create a new Unique Job strategy that we can later on apply. Apart from rescheduling, our strategy won't differ from the WhileExecuting, so we can use it as a base.

module SidekiqUniqueJobs
  module Lock
    class WhileExecutingReschedule < WhileExecuting
      MUTEX = Mutex.new

      def synchronize

        if (@locked = locked?)
          # We use sleep just to prevent from a pointless, extremely fast looping
          # in case all the jobs have the same unique key
          sleep 0.1
      rescue Sidekiq::Shutdown
        logger.fatal { "the unique_key: #{@unique_digest} needs to be unlocked manually" }
        # If we were able to obtain lock, we need to release it after processing
        if @locked
          SidekiqUniqueJobs.connection(@redis_pool) { |conn| conn.del @unique_digest }


Applying this strategy is super easy. We just need to replace while_executing strategy with while_executing_reschedule:

class ApplicationWorker
  include Sidekiq::Worker

  # sidekiq_options unique: :while_executing
  sidekiq_options unique: :while_executing_reschedule

Cover photo by: Alexandre Duret-Lutz on Creative Commons 2.0 license. Changes made: added an overlay layer with an article title on top of the original picture.

Copyright © 2022 Closer to Code

Theme by Anders NorenUp ↑