Mike Perham

Currently we have a simple counter metric being fired at the end of every callback so you can monitor `sidekiq.batches.success` for instance. As they finish you'll see blips on a...

@michelle-settle Please don't poke me for updates. It is in my backlog.

`perform_inline` has its own implementation which skips S::Client. https://github.com/sidekiq/sidekiq/blob/4d805754d82f33522db68e5c6dee22b0e73a7ef5/lib/sidekiq/job.rb#L215 so this line is missing: https://github.com/sidekiq/sidekiq/blob/4d805754d82f33522db68e5c6dee22b0e73a7ef5/lib/sidekiq/client.rb#L107

Perhaps `S::Testing.inline` and `perform_inline` should use the same implementation.

Could you be more specific? What's the problem?

Oh, you are migrating from systemd to docker. I don't have a good answer for you. I can't help with 3rd party tooling which I have no experience with.

I don't think we have a Docker wiki page but we do have https://github.com/sidekiq/sidekiq/wiki/Kubernetes

And starting sidekiqswarm in Docker should be very similar to Sidekiq, just replace `bundle exec sidekiq` with `SIDEKIQ_COUNT=2 bundle exec sidekiqswarm`. Make sure you set SIDEKIQ_COUNT to reflect the CPU...

You'd have to look up the associated PR which added that feature. I don't recall if there was any logical need or if this was just an oversight.

Ok, it sounds like skipping death handlers is a no-no. We can't allow an option which breaks functionality.