Request limits exceed container limits
An AI wrote this pod spec for a JVM service when you asked it to "add resource limits so it doesn't hog the node." It deploys, but the service gets OOMKilled under load and the node also gets noisy-neighbor CPU stalls. What's misconfigured?
Implement
pod_resource_verdict(cpu_request: str, cpu_limit: str, memory_request: str, memory_limit: str) → strExamples
in
["2","500m","256Mi","4Gi"]out"invalid_cpu"in
["500m","500m","2Gi","2Gi"]out"guaranteed"in
["500m","1","256Mi","4Gi"]out"oomkill_risk"What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 16 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.