Troubleshooting Clusters and Nodes

A node flips to NotReady at 2 AM. Pods start getting evicted. Alerts fire. You get paged. Where do you start?

This is the most common troubleshooting scenario in Kubernetes, and it maps directly to the CKA exam. Troubleshooting accounts for 30% of the exam weight, the largest single domain, and "troubleshoot clusters and nodes" is the first competency listed. The exam gives you a broken cluster and expects you to fix it under time pressure. This lesson walks through a systematic diagnostic workflow that covers the most common causes and points you to the right next step.