Track all workers concurrency jobs

What does this MR do and why?

Before this, we only track the number of concurrent jobs when concurrency_limit attribute is defined on the worker class.

Now, we track all worker classes concurrent jobs in order to support the work for circuit breaking/throttling all sidekiq workers based on database usage.

This is primarily done by setting concurrency_limit to 0 by default for all workers.

This change is also behind track_sidekiq_concurrency_limit_execution feature flag.

References

gitlab-com/gl-infra/data-access/durability/team#135

What is concurrency limit in sidekiq - https://docs.gitlab.com/development/sidekiq/worker_attributes/#concurrency-limit

Screenshots or screen recordings

Before	After

How to set up and validate locally

1. Enable the feature flag track_sidekiq_concurrency_limit_execution

2. Restart sidekiq

3. In rails console:

   5.times { Chaos::SleepWorker.perform_async(60) }
   Gitlab::SidekiqMiddleware::ConcurrencyLimit::ConcurrencyLimitService.concurrent_worker_count(Chaos::SleepWorker.name)

4. The above should return 5

5. Metric should be emitted too. For some reason, I don't see the metric anymore in my GDK, but it appeared previously. Please help me to confirm this.

   curl -s 'gdk.test:3807/metrics' | rg sidekiq_concurrency_limit_current_concurrent_jobs_total

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.