This is why production services should have health checks that actually verify whether the service is doing its job in the system, not whether it's merely runni

This is why production services should have health checks that actually verify whether the service is doing its job in the system, not whether it's merely running the process. A process can be completely broken while still technically running, and vice versa (at least with janky ps-parsing scripts).