[While doing some testing today I noticed that Sidekiq Cluster will only use one CPU by default](https://gitlab.com/gitlab-org/omnibus-gitlab/-/issues/5797#note_552414062) as it only has one "worker" (set via `queue_groups`). As such I think we should have Omnibus [do something similar like it already does for Puma](https://gitlab.com/gitlab-org/omnibus-gitlab/-/blob/master/files/gitlab-cookbooks/gitlab/libraries/puma.rb#L39) and for it to scale Clusters process count higher based on the same number of CPUs to ensure there's no performance bottlenecks though this should only happen when Sidekiq is standalone and not being run with other components.