Kubernetes · Verify

Kubernetes Rollout & Failure Check

Verify the selected workload’s rendered contract, availability or state recovery during its required lifecycle transitions.

MCP library 1.6.1 · Instruction library 1.6.1 · English technical instructions

← Back to search

Use this workflow

Connect Vibe Coding Library to a coding agent that supports MCP over Streamable HTTP. Choose Other / MCP agent for a neutral prompt, or a ready-made Codex or Claude Code invocation. Add your task details below.

Goal

Verify the selected workload’s rendered contract, availability or state recovery during its required lifecycle transitions.

Context

Identify target overlay/chart and cluster version, workload identity, probe/termination settings, availability budget, stateful dependencies and authorized test environment.

Scenarios

Select the cases that test the requested guarantee. Define expected state/effects before execution; use the authorized environment and a valid control.

  • Render and validate the exact overlay; reject selector, port, reference, schema, or policy mismatches.
  • Hold a dependency unavailable during startup and verify startup/readiness behavior without liveness restart amplification.
  • Send traffic during rollout and SIGTERM; verify drain, availability budget, progress deadline, and rollback signal.
  • Test voluntary eviction through its supported mechanism separately from involuntary pod/node loss; inspect the relevant budget, rescheduling, volume attach and state recovery.
  • Drive representative saturation and verify requests, limits, HPA signal, stabilization, backlog, and overload response.

Verification

Compare the rendered resources, readiness/endpoints, accepted traffic, rollout progress, drained work and state recovery against the selected workload promise.

Execute and decide

Record the exact rendered artifact, target API version, namespace/workload identity and expected availability/drain/state criteria. Validate selectors, ports and references, then establish healthy endpoints and a normal request/state control in the authorized cluster when runtime evidence is required.

Introduce only the selected startup, dependency, termination, rollout or disruption condition. Observe actual accepted traffic, readiness/endpoints, progress/events, unfinished work and persisted state through recovery. PDB behavior applies to the relevant eviction path; it does not prevent involuntary loss or replace a workload controller's rollout policy.

Return passed, failed or blocked for the stated transition and environment. Feed a failure to the owning chart/operator/resource repair, repeat its transition/control after correction and finish Patch Checking for source changes. Without a suitable cluster, report the render/policy checks actually performed and the missing runtime gate.

Workflow ID: kubernetes-rollout-failure-check · View the versioned source · Shared workflow and authority rules