Skip to main content

Rollout, disruption and recovery evidence — assessment

Score 0 absent/incorrect, 1 prompted/incomplete, 2 independently supported with explicit limits, 3 justified fresh transfer. Every category requires at least 2; a high average cannot hide a missing category.

Required evidence​

  • KO1 — Probe contract: Show three distinct failures and observed consequences; separate readiness from liveness.
  • KO2 — Disruption: Distinguish eviction, rollout and involuntary loss, including a constrained scheduler case.
  • KO3 — Work integrity: Preserve the worker effect/acknowledgement invariant across termination and retry.
  • KO4 — Diagnosis: Retain original Pending, crash-loop, image-pull and storage failures with causal evidence before repair.
  • KO5 — Security: Retain actual RBAC/non-root/capability/image checks and positive/negative traffic tests; identify enforcement prerequisites.
  • KO6 — Recovery: Retain manifests, revision/digest, client observations, rollback, post-recovery data checks and scoped teardown.

Calibration​

At three healthy replicas with integer minAvailable 2, one healthy eviction may be permitted; this is not a guarantee against node loss or a Deployment rollout. A stalled rollout with a Pending Pod is not repaired by weakening a readiness predicate. Successful rollback of images does not restore a database schema. Report missing cluster execution as pending.

Delayed transfer and integrity​

After at least 48 hours, have a reviewer change one assumption or select a previously unseen failure. Predict the outcome, collect fresh evidence and explain discrepancies without the calibration text. Preserve mistakes, repairs, collaboration and generated-assistance disclosures. Record unavailable tools or reviewers as pending evidence, never fabricated completion. Author reference checks are not learner runtime experiments.