Catched a deadlock on Sidekiq 7+ set up using middleware.

[akapitula]

Describe the bug In my configuration the deadlock was not supposed to happen but still was happening, the gem was configured using sidekiq middleware. sidekiq (7.2.1) sidekiq-unique-jobs (8.0.9)

Expected behavior Not to lock the job. But it still was locked for hours, untill i've resetted the redis.

Current behavior

2024-02-29T08:46:04.343Z pid=23072 tid=cng uniquejobs=client until_executed=uniquejobs:a95805955a77049ece74f49084024c62 INFO: Skipping job with id (281f76da6925b8a897df39d1) because lock_digest: (uniquejobs:a95805955a77049ece74f49084024c62) already exists

Worker class

class PendingPropositionWorker
  include Sidekiq::Worker
  sidekiq_options lock: :until_executed, on_conflict: :log

  def perform
    PropositionProcessService.call
  end
end

sidekiq.rb initializer:

require "sidekiq-unique-jobs"

redis_config = { url: ENV['REDIS_URL'], network_timeout: 5, pool_timeout: 5 }

Sidekiq.default_job_options = { backtrace: true, retry: 5 }

Sidekiq.configure_server do |config|
  config.redis = redis_config
  config.error_handlers << Proc.new { |ex, ctx_hash| Rollbar.error(ex, params: ctx_hash)}
  config.client_middleware do |chain|
    chain.add SidekiqUniqueJobs::Middleware::Client
  end

  config.server_middleware do |chain|
    chain.add SidekiqUniqueJobs::Middleware::Server
  end

  SidekiqUniqueJobs::Server.configure(config)
end

Sidekiq.configure_client do |config|
  config.redis = redis_config
  config.client_middleware do |chain|
    chain.add SidekiqUniqueJobs::Middleware::Client
  end
end

Additional context In the sidekiq interface the lock was not shown