FREE KUBERNETES BREAK/FIX
Three broken clusters.
No signup.
Apply the bad config to a disposable cluster, inspect what Kubernetes is actually telling you, make a theory, then fix it. The answers are here too. No crippled-demo nonsense.
The Service Has No Endpoints
The pods are healthy. The Service exists. Traffic goes nowhere. Figure out why before blaming DNS.
LAB 02 · EASY · 5–10 MINPending Forever
There is node capacity, but the workload will not schedule. Read the scheduler instead of guessing.
LAB 03 · MEDIUM · 10–15 MINThe PVC Never Binds
A PV exists with enough capacity. The claim still sits Pending. Storage has more predicates than “disk exists.”
HOW TO USE THESE
Break it on purpose.
Use a disposable local or lab cluster. Apply the supplied manifest. Do not read the solution until you can explain the symptom in your own words. The point is not to memorize a command; it is to build the habit of turning cluster state into a useful theory.
Comfortable in a shell but not sure where this fits into a career move? Start with Help Desk → Cloud Infrastructure: what actually matters.
IF THESE HELP
There are seven more.
The full pack adds NetworkPolicy, probes, ConfigMaps, RBAC, CronJobs, ResourceQuota, and startup/readiness interactions, with the same break → inspect → explain → fix format.