bust-a-kube

This repo houses various sets of problematic k8s resources, can you debug them?

Each "bug" lives in it's own subdirectory. The subdirectories are sorted into the following categories:

• workload - Covering containers, pods, and all controller types in Kubernetes
• services-networking - Covering services, discovery (DNS), network policies, and ingress
• config-storage - Covering configMaps, secrets, the downward API, and the elements of the persistent volume subsystem
• observability - Covering workload scaling, metrics, logs, and events
• cluster - Covering control plane components, kubelets, various plugins (Network, Storage and Container Runtimes)
• security - Covering Role-based access control, security contexts, ServiceAccounts, and various policies

The subdirectories contain the following files:

• README.md - explains the context of the problem the resources exhibit
• problem.yaml - the bug, apply this manifest to your cluster and see if you can debug it
• solution.md - an explanation of what is wrong, how to diagnose such a problem and how to fix it

Each problem is designed to be independent of one another and can be completed in any order.

WARNING: Do not run these manifests on a cluster you care about!!

These problem resources can impact the functioning of you cluster and/or it's applications (they are problems after all!). All of the problems are designed to be compatible with recent versions of Kubernetes (1.18+). Some developer oriented tools, such as Docker Desktop, minikube, kind, etc. do not provide a complete implementation of Kubernetes with their default configuration and therefore some bust-a-kube challenges may not work. To stand up a suitable single node Kubernetes test cluster, you can run the RX-M k8s.sh script on a plain vanilla Ubuntu x64 VM.