IaaSOKE-bootcamp
Deploy, monitor, and scale an application on a Luna-provisioned Oracle Kubernetes Engine (OKE) cluster. OKE is OCI's managed Kubernetes service. You will use Helm, a package manager for Kubernetes, to install the application and monitoring tools, then observe how the application behaves as traffic and replica counts change. A customized web application with a public endpoint that stays the same as application pods are added or replaced. Proxies beside your application containers that manage service traffic and report request metrics. Istio base registers the Kubernetes resource types it needs; istiod configures the proxies. Collects and stores metrics from Istio so you can examine request rates, errors, and latency. It supplies the data queried by Kiali and Grafana. Displays a graph of service communication and helps you inspect traffic, request errors, and workload health. Displays metrics over time in a prepared dashboard so you can compare baseline traffic, a load burst, and scaling activity. Adjusts the number of application pods according to CPU demand. It scales application replicas, not worker nodes. Customize application settings and install or upgrade a Helm release. Explain how a Kubernetes Service keeps the application reachable as its pods change. Follow service traffic in Kiali and interpret request rates, success rates, and latency in Grafana. Scale replicas manually, then observe CPU-based autoscaling as load rises and falls. Distinguish readiness from liveness and use pod status and events to investigate health warnings. Optionally, replace one application pod and observe Kubernetes restore the desired replica count. Have these ready before the hands-on portion: Access to your assigned Luna desktop and the lab repository or prepared checkout. A working kubeconfig and your assigned cluster context, provided by the instructor. A Bash terminal with OCI CLI, kubectl, Helm, Git, and curl installed. Basic familiarity with terminal commands and editing a YAML file. No application coding or container-image build is required. The cluster already includes. , which supplies CPU readings for this lab's HPA and CPU/memory readings for kubectl top, and. , which manages TLS certificates and is a dependency of the OCI-managed Metrics Server add-on. You do not reinstall these components. Prometheus supplies dashboard metrics; Metrics Server supplies this HPA's CPU input. 60 minutes of hands-on exercises and debrief. The workshop allocates an additional 30 minutes for the introductory lecture while the cluster provisions, for 90 minutes total. The cluster must be ready before hands-on work starts. Use only your assigned training cluster. Students do not manually destroy infrastructure; Luna starts automated cleanup when the session ends or expires. Follow the lab instructions to stop the CPU load after the scaling exercise.