Rendimiento y fiabilidad

IO/CPU limits, restarts, healthchecks, reliability patterns.

cgroups limits healthcheck restart policy observability performance