Debugging KCNA Practice Questions (Page 5)

Questions 21–23

1. 21application · hard

   A user reports that a web application returns a 502 Bad Gateway error. The application runs in Kubernetes behind an Ingress. The Ingress controller logs show that it cannot connect to the backend Service. The Pods are running and the Service endpoints are populated. What should the team check next?

   • ACheck the application's liveness probe configuration.
   • BCheck the Service's `externalTrafficPolicy` setting.
   • CCheck the network policy that may be blocking traffic from the Ingress controller to the Pods.
   • DCheck the Ingress controller's replica count.
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   Which approach is most effective for debugging a distributed cloud native application?

   • AUse centralized logging, metrics, and tracing to correlate behavior across services
   • BAttach a debugger to a single running container and step through code
   • CRestart all pods until the issue disappears
   • DInspect the source code of every microservice manually
   Reveal Solution Discussion Select an answer first
3. 23expert · hard

   A company runs a critical payment service in Kubernetes. The service has a liveness probe that checks `/healthz`, which returns 200 only when the service can reach its database. Recently, the database had a brief outage. During the outage, the service's Pods were restarted repeatedly, causing a cascading failure. After the database recovered, the service took several minutes to become healthy again. What is the most likely cause of the cascading failure?

   • AThe Service's `externalTrafficPolicy` was set to `Local`, causing the Ingress to route traffic to unhealthy Pods.
   • BThe HorizontalPodAutoscaler scaled down the Deployment because the database was unavailable.
   • CThe readiness probe failed, so the Service removed the Pods from endpoints, but the Deployment controller scaled down the replicas.
   • DThe liveness probe failed during the database outage, causing the kubelet to restart the containers.
   Reveal Solution Discussion Select an answer first