The Kubernetes API is the cluster’s RESTful interface for accessing resources and communicating with cluster components. The API server exposes it from the control plane. Users and tools reach the cluster through this entry point; worker nodes receive instructions through the control plane and report their status back. section 1.5 section 1.4.1
Resources such as Deployments are API objects with resource URIs. A Deployment comprises ReplicaSets, which run pods. Think of an API request as acting on a resource, rather than directly starting a container on a node. section 1.4.2 section 1.4.1
After authentication, HTTP methods express the requested operation:
- GET retrieves resource data; POST creates a resource. section 1.4.2
- PATCH updates part of a resource; PUT replaces it; DELETE removes it. section 1.4.2
apiVersion: v1
kind: Pod
metadata:
labels:
run: nginx
name: nginx
spec:
containers:
- image: nginx
name: nginx
This Pod manifest shows the resource fields supplied in YAML: apiVersion, kind, metadata, and spec. It describes the requested pod before a command sends it to the cluster. section 1.9
Create and check a resource
Run `kubectl create -f my-pod-manifest.yaml`; the shown result is `pod/nginx created`. Use `kubectl get po` to inspect pods afterward rather than treating the creation message as the only check. For API access, also account for authentication and RBAC permissions: users and service accounts may be allowed different operations. section 1.9 section 2.1.2 section 9.3.2
