etcd Troubleshooting and Cluster Recovery

Your kube-apiserver is returning 500 errors. Pods aren't scheduling. kubectl get nodes hangs. Before you start debugging the scheduler or kubelet, check etcd. If etcd is unhealthy, nothing in the cluster works because every API call reads from or writes to it.

etcd troubleshooting is a core CKA skill in the Troubleshooting domain (30% of the exam). You need to know how to diagnose health problems, manage cluster membership, handle space quota alarms, and reclaim disk space through defragmentation. This article covers the operational commands you'll reach for when etcd is misbehaving. For backup and restore procedures, see etcd Backup, Restore, and Disaster Recovery.