While resources such as replicasets, replicacontroller, services are useful to manage a desired state of the application using pods, they are tied to a particular version of the application. To rollout the new version of the application, these resources are not helpful. The deployment object helps to control and release the release of new versions of the application.

Deployments helps to move from one version of application to another using a user-configurable rollout process. It also uses health checks to ensure that new version of the application is operating correctly and enable additional options such as stopping the deployment if too many failures occur.

In Kubernetes, pods are the smallest unit of deployment for compute. A pod can contain one or more containers for the application. However, you would often want one or more instances of pods running in parallel for reasons like scalability, sharding or other reasons. A ReplicaSet ensures that a specified number of pod replicas (one or more) are running at any given time.

ReplicaSets are the building blocks used for the higher level concepts such as deployments or horizontal pod autoscalers. By ensuring that specified number of pod replicas are running, they provide self-healing for applications for certain failure conditions such as node failures or network partitions. Most of the time, one would be using higher level concepts like deployments instead of directly using ReplicaSets.