CKA Course Notes

The exam tests the foundational elements of a Kubernetes cluster and the procedures for maintaining its stability. Understanding cluster architecture, control plane components, and node management is critical for administering a Kubernetes environment.

• Understand the distinction between cordoning and draining a node. Cordoning *prevents* new Pods, while draining *removes* existing ones.

• Prioritize understanding the core control plane components and their responsibilities. A strong foundation here will help you answer many questions.

The exam tests the core components responsible for cluster operation and the tools used to establish a Kubernetes control plane. Understanding these elements is critical for troubleshooting cluster behavior, scaling deployments, and ensuring high availability. The exam will test your ability to diagnose issues related to control plane functionality and node agent behavior, as well as your understanding of the kubeadm bootstrapping process.

• The kubelet manages Pods, while the container runtime executes the containers *within* those Pods. They are distinct but interdependent.

• Focus on understanding the *responsibilities* of each component, not just their definitions. The exam emphasizes practical application.

The exam tests the foundational security and governance aspects of Kubernetes clusters, particularly those managed with kubeadm. Understanding how to securely join nodes, manage access with kubeconfig, and define permissions with RBAC is critical for maintaining cluster integrity and controlling user access.

• A Role is namespace-scoped, limiting its permissions to resources within a single namespace. A ClusterRole is cluster-scoped and can be applied to the entire cluster or bound to specific namespaces.

• Always consider the scope of permissions required when choosing between a Role and a ClusterRole. Favor Roles for namespace-specific access.

Kubernetes lifecycle management involves controlling access, resource usage, and identity within your cluster. The exam tests how namespaces, RBAC, ServiceAccounts, and ResourceQuotas work together to create a secure and organized environment.

• **RoleBinding vs. ClusterRoleBinding:** RoleBindings are namespace-scoped, while ClusterRoleBindings apply cluster-wide. Choose the narrowest scope possible.

• Always consider the scope of permissions when choosing between RoleBindings and ClusterRoleBindings. Favor namespace-scoped solutions whenever possible.

Workload orchestration defines how Kubernetes manages your applications. The exam tests the core controllers – Deployments, StatefulSets, DaemonSets, Jobs, and CronJobs – and how they handle different application needs. Understanding these controllers is critical for deploying and managing applications reliably and efficiently.

• **Deployments vs. StatefulSets:** Deployments are for stateless applications; StatefulSets are for stateful applications requiring persistent identity and ordered operations.

• Carefully analyze the workload requirements before selecting a controller. Consider whether the application is stateful or stateless, and whether it requires persistent identity.

Kubernetes Services provide a stable IP address and DNS name for a set of Pods, abstracting away the underlying Pod changes. Understanding Service types and their behavior is critical for exposing applications within and outside the cluster. The exam tests the different Service types, how they function, and how to troubleshoot common issues.

• **ClusterIP vs. NodePort:** ClusterIP is internal-only, while NodePort exposes the Service on each node's IP.

• Carefully read the requirements of each scenario. Look for keywords like 'internal-only,' 'external access,' or 'direct endpoint discovery.'

The exam tests Kubernetes networking, a critical aspect of cluster operation. It covers how external traffic reaches your applications, how services are discovered within the cluster, and how to control network traffic flow for security and isolation.

• **Ingress vs. Service:** Ingress manages external access, while Services provide internal load balancing and discovery.

• Pay close attention to the wording of scenario questions, especially keywords like 'external access,' 'internal discovery,' and 'traffic control.'

Storage management in Kubernetes involves provisioning, attaching, and managing persistent storage for your applications. The exam tests how Kubernetes handles persistent volumes, claims, and storage classes, enabling applications to reliably store and retrieve data.

• **PersistentVolume vs. PersistentVolumeClaim:** PV represents the storage resource; PVC represents the request for storage.

• Carefully read the scenario and identify the specific storage requirements (e.g., read-write access, data persistence).

Troubleshooting is a critical CKA skill. The exam tests diagnosing and resolving common issues encountered in a Kubernetes cluster, from Pod scheduling problems to Service connectivity failures and control plane instability.

• Distinguish between `ImagePullBackOff` and `CrashLoopBackOff`: one is an image retrieval problem, the other a container runtime failure.

• Always start with `kubectl describe` to gather information about the failing object and its events.

The exam tests troubleshooting common issues within a Kubernetes cluster, emphasizing practical skills and diagnostic techniques. Effective troubleshooting is critical for maintaining cluster stability and application availability, and is a significant portion of the CKA exam.

• **kubelet vs. Container Runtime:** The kubelet manages Pods and reports status; the container runtime (e.g., containerd, CRI-O) actually runs the containers.

• Prioritize troubleshooting steps. Start with the most likely cause and work your way down.