Module 15 case studies
Module 15 · Put your knowledge to work in context.
Module 15 is a multiple-choice examination organised around five domains: Fundamentals and Core Concepts (20%), Planning and Design (20%), Implementation and Configuration (25%), Operations, Monitoring, and Maintenance (20%), and Troubleshooting and Governance (15%). The weighting tells you where the marks sit: implementation and configuration is the single largest block, and the two 20% planning and operations domains together account for a further 40%, so an approach that only drills definitions and terminology will leave a large share of the paper untested. The Module 15 case studies are 5 cases and about 3,979 words.
1Nightly Evictions in a Cluster That Is Only 71% Committed
A 9-node cluster (each 8 vCPU / 32 GiB, ~31 GiB allocatable) runs two workloads. The batch namespace holds 40 CronJob-spawned pods that declare no CPU request and no memory request, only limits.memory: 512Mi ; measured peak memory per pod is about 900 MiB.
796 words, about 4 min2Pod CIDR Exhaustion on a VPC-Native Cluster
A production Kubernetes cluster runs VPC-native (alias IP) networking. The cluster's pod secondary range is 10.16.0.0/21 , and every node is allocated a /24 from it.
741 words, about 4 min3Rolling Updates That Pass Readiness and Still Time Out
A web Deployment runs 24 replicas across three zones with rollingUpdate: maxSurge: 25%, maxUnavailable: 0 , and an HPA with minReplicas: 24 , maxReplicas: 40 , target CPU 70%. The readiness probe is GET /healthz on port 8080, periodSeconds: 5 , initialDelaySeconds: 0 , default failureThreshold .
818 words, about 4 min4Prometheus at 800k Series With a Noisy Pager
A single Prometheus server scrapes 400 targets every 15 seconds and holds about 800,000 active series against a 500 GiB volume with 30-day retention. A platform-owned exporter exposes request duration seconds with a user id label carrying roughly 120,000 distinct values and an http path label that includes raw IDs.
779 words, about 4 min5A Failing Admission Webhook and a Denied Egress Path
A shared cluster hosts 14 tenant namespaces. A validating admission webhook backed by a single OPA Gatekeeper replica enforces an image-registry policy with failurePolicy: Fail , timeoutSeconds: 3 , and a namespaceSelector that matches all namespaces except kube-system .
845 words, about 4 min