CKA v1.35 · Troubleshooting (30%) + Servicing and Networking (20%) → "Troubleshoot services and networking" / "Understand and use CoreDNS" Target time: 10 minutes
Cluster-wide DNS has stopped working. Pods can reach each other by IP but
every name lookup fails — nginx.default.svc.cluster.local, external names,
everything.
Find the cause and restore DNS. Do not delete the CoreDNS Deployment; repair the actual fault.
Confirm with a lookup from a fresh pod when you're done.
NOTE: this scenario breaks DNS for the WHOLE cluster. If you abandon it mid-drill, other scenarios will fail in confusing ways. Repair with: bin/reset scenarios/cka/06-coredns-broken bin/doctor --fix