Introduction to Kubernetes

You will learn where Kubernetes comes from, what problems it solves, and how the architecture of Control Plane and Worker Nodes is structured. By the end of this block, you will understand why Kubernetes has become the standard for container orchestration.

You will compare different Kubernetes distributions (managed services like EKS, AKS, GKE vs. self-hosted options like k3s, kubeadm) and evaluate which approach fits which scenario. Afterwards, you can choose the right Kubernetes variant for your project.

You will set up kubectl, connect to your cluster, and run your first commands. By the end, you will navigate resources confidently, read logs, and understand the output of kubectl get, describe, and explain.

You will install Krew as a plugin manager and extend kubectl with useful plugins. Then you will set up K9s, a terminal UI that lets you visually monitor your cluster, debug Pods, and manage resources in real time.

You will create your first Pod, observe its lifecycle, and understand why Pods are the smallest deployable unit in Kubernetes. Then you will deploy multi-container Pods and learn when the sidecar pattern makes sense, including the native sidecar support available since Kubernetes 1.33.

You will configure LivenessProbes, ReadinessProbes, and StartupProbes on a running Pod. Then you will simulate a failure and watch how Kubernetes automatically restarts the unhealthy container. This gives you a hands-on understanding of self-healing in practice.

You will create your first Service and use it to connect two Pods. Along the way, you will understand how Kubernetes networking works and how Services act as a stable endpoint in front of changing Pod IPs.

You will configure ClusterIP, NodePort, and LoadBalancer Services and see live how network access to your Pods changes. You will learn about Traffic Distribution (Kubernetes 1.34) to keep traffic preferentially in the same zone or on the same node. Afterwards, you will know which Service type to use in which scenario.

You will distribute an application across multiple Nodes, simulate a Node failure, and watch how Kubernetes automatically reschedules the Pods. This way, you experience high availability directly on a running cluster.

You will create ReplicaSets, scale the replica count up and down, and watch how Kubernetes ensures the desired number of Pods. Using In-Place Pod Resize (Kubernetes 1.35 GA), you adjust CPU and memory of running pods without restarting them. Along the way, you will learn why ReplicationControllers are deprecated and how ReplicaSets replaced them.

You will create your first Deployment, update the container image, and watch how Kubernetes performs a rolling update. Afterwards, you will understand why Deployments are the central object for running your applications.

You will compare Rolling Update, Recreate, Canary, and Blue-Green Deployments and execute each strategy yourself. By the end, you will know which strategy suits which risk profile and how to trigger rollbacks when issues arise.

You will create Namespaces, move resources between them, and set Resource Quotas to limit CPU and memory per Namespace. This way, you will understand how teams cleanly separate their workloads on a shared cluster.

You will access Services via DNS names, test name resolution across Namespaces, and debug typical DNS issues. Afterwards, you can reliably reach any application in the cluster by its Service name.

Kubernetes separates orchestration from runtime, networking, and storage through the standard CRI, CNI, and CSI interfaces. Using one example, you trace the components involved when a Pod starts, receives network connectivity, and when a PersistentVolumeClaim triggers provisioning through the prepared CSI driver; you then place containerd, CRI-O, Calico, and Cilium in that model. This lets you assign later failures to the correct layer.

You will create PersistentVolumes and PersistentVolumeClaims, attach them to a Pod, and verify that data survives a Pod restart. This takes you from theory to a working storage configuration in Kubernetes.

You will configure StorageClasses and see how dynamic provisioning automatically creates volumes. Using VolumeAttributesClass (Kubernetes 1.34 GA), you modify performance parameters (IOPS, throughput) of existing volumes without recreation. Along the way, you will learn how CSI drivers connect different storage backends and how to choose the right StorageClass for your workload.

You will store configuration data in ConfigMaps and sensitive values in Secrets, and mount both as environment variables or volume mounts into your Pods. By the end, you will cleanly separate configuration from the container image, just as it should be in production.