Pods

A pod is the minimum computing unit that can be run in a Kubernetes cluster.

A pod runs one or more application containers:

• Typically, a pod consists of a single container, being a wrapper on top of it. This is the most common scenario for using pods in a Kubernetes cluster.

• A rarer case: a pod can consist of multiple containers. These containers can communicate with each other and share resources assigned to the pod (e.g., storage).

From a Kubernetes perspective, one pod is one application instance. If you want to have multiple instances of an application for high availability, use workload controllers, which allow you to replicate pods and control their lifecycle. An example of such a controller is Deployment.