Pods
In depthPlatform
Also known as: Kubernetes Pods
The smallest unit in Kubernetes: a group of one or more containers that run together.
Pods, explained
Written by EduVerseWhat it is
A pod is the smallest thing Kubernetes runs. It wraps one container, or sometimes a few closely related ones, that share the same network address and can share storage. Kubernetes doesn’t start containers directly: it places pods on machines in the cluster and keeps track of whether they’re healthy.
Why teams use it
Grouping containers into pods gives Kubernetes one unit to place, restart and replace. Pods are meant to be temporary: when one crashes or a node goes down, a new pod with a new name and IP address takes its place. That’s why you put a Service in front of them instead of connecting to a pod directly.
An example from work
You run kubectl get pods and see orders-api-7d9f8c6b5-x2k4q with status CrashLoopBackOff and a restart count that keeps climbing. Your container starts, crashes, and Kubernetes keeps retrying with longer pauses in between. Running kubectl logs with --previous on that pod shows the output of the last crashed run, which usually contains the real error.
Our own explanation, not a quote from the book.
In the book
Sentences from The Software Realm, Decoded that mention Pods, exactly as printed.
12 more passages about Pods in the full book
Read every conversation where Pods comes up, with the interactive slides and demos.
Where it fits
Running containers inside Kubernetes
Coverage in the book
Covered in depth: multiple pages with explanations, examples and simulations.
