School-district clients time out connecting to a ClusterIP Service. Which objects should the administrator inspect to see which Pod IPs currently back the Service?
Select an answer to reveal the explanation.
Short Explanation
When the Service phone line is dead, open the Endpoints/EndpointSlices directory to see which Pod IPs are actually listed.
Full Explanation
School-district clients timing out connecting to a ClusterIP Service should lead the administrator to inspect Endpoints or EndpointSlices associated with the Service, which list the ready Pod IPs and ports selected as backends. Those objects are the control-plane view of who should receive traffic; empty or wrong backends explain many connect timeouts before deeper dataplane debugging.
ResourceQuota status does not enumerate Pod IPs bound to each Service name. CertificateSigningRequest lists do not publish backend IPs after kubelet rotation for this purpose. PersistentVolume claims named after the Service do not store the last known Endpoints.
Exam caveat: EndpointSlices are the scalable modern API; Endpoints may still appear depending on cluster version and tooling. Operational check: kubectl get endpoints,endpointslice -l kubernetes.io/service-name=<svc>, verify Pod readiness and selectors, then test from a district client Pod. This distinction matters on the CKA because distractors often mix adjacent objects that sound related but solve a different control-plane problem. This distinction matters on the CKA because distractors often mix adjacent objects that sound related but solve a different control-plane problem.