concept Node in category kubernetes

This is an excerpt from Manning's book Core Kubernetes MEAP V02. Login to get full access to this book.

The Node is the base unit of compute in a Kubernetes cluster. A Node usually consists of the following components:

The following diagram displays what components form a Node.

Figure 1.3. A Node

Every server in a Kubernetes cluster has a kubelet binary running on the Node, and typically another application called kube-proxy. The kubelet binary is in charge of containers and the Pod lifecycle: start, stop, restart, delete, etc. It also handles various Node level operations. Some installers run kubelet as a container, and some do not.

# Get the manifest for the project
curl -LO https://raw.githubusercontent.com/lionkube/core-kubernetes/master/code/chapter-09/pod-with-resources.yml

# Create a single web server Pod
# Note that we’ve added `resources` field to the pod declaration
# This allows us to set limits for memory and cpu usage of the Pod
kubectl apply -f pod-with-resources.yml

# Wait a few seconds for the Pod to finish starting up

# Here we’re going to fetch the pod and container ID from Kubernetes before moving on
# We’ll use these when we start looking at cgroup information in the Node

# This retrieves the ID of the pod in Kubernetes
# Store this value to a local variable pod_id using: export pod_id=...
kubectl get pods httpd-with-resources -o=jsonpath='{.metadata.uid}'

# This retrieves the ID of the container being run in the Pod
# Strip off the leading containerd:// and store the value to a local variable container_id using: export container_id=...
kubectl get pods httpd-with-resources -o=jsonpath='{.status.containerStatuses[0].containerID}'

# Make sure to leave this shell open for the next steps in the lab!
# We need the values stored in those variables and don’t want to lose them

copy