With all of these types of service discoveries available, we should be able to handle any container scaling that we want, but we need to keep something very important in mind: almost all service discovery tooling is intimately bound to the system used for deploying and managing the containers (also known as container orchestration) due to the fact that updates to container endpoints are generally just an orchestration system implementation detail. Because of this, service discovery systems usually aren't as portable as one might like, so the choice of this infrastructure piece usually gets decided by your orchestration tooling (with a few exceptions here and there).