Research Pod Resource Limits for Sidekiq
Based on the output of #1990 (comment 350329568) ; we need to investigate how to implement pod resource limits for Sidekiq.
According to Sidekiq Memory Killer docs, it'll run as legacy mode by default. Meaning that:
In _legacy_ mode, if the Sidekiq process exceeds the allowed RSS then an irreversible delayed graceful restart will be triggered. The restart of Sidekiq will happen after `SIDEKIQ_MEMORY_KILLER_GRACE_TIME` seconds.
So for K8s envs (Helm charts), ideally we should configure that to be handled per container / pod and disable the lib by default.