Ordered Creation, Update and Deletion

Because StatefulSet members often have real dependencies on each other - a follower may need its leader to already exist, for instance - StatefulSets, by default, create and start Pods one at a time in strict ordinal order (podManagementPolicy: OrderedReady), only moving on to the next ordinal once the current one reports ready; a Parallel setting exists to relax this when your application genuinely has no such startup dependency. Updates follow their own careful, opposite-direction order by default: a StatefulSet rolling update proceeds from the highest ordinal downward, and a partition value lets you stage a rollout by only updating ordinals at or above that number, deliberately leaving lower ordinals untouched - useful for testing a change on a subset of replicas before committing further. When comparing a StatefulSet's currentRevision (what is actually running) against its updateRevision (what it is being asked to converge to) alongside individual Pod readiness, you can see exactly how far a rollout has progressed and where it has stalled. Because of the strict ordering, one broken Pod can genuinely deadlock the whole rollout - which is precisely why understanding your specific application's membership and bootstrap requirements (does it truly need this ordering, or was that assumed unnecessarily) is part of operating this controller correctly.

Analogy: Kubernetes is a thermostat, not a remote control. You declare the temperature; independent controllers measure reality and keep acting. Debugging means finding which sensor, rule or actuator prevents convergence.

Read the object as evidence

kubectl get RESOURCE NAME -n NAMESPACE -o yaml
kubectl describe RESOURCE NAME -n NAMESPACE
kubectl get events -n NAMESPACE --sort-by=.metadata.creationTimestamp
kubectl explain RESOURCE.spec

Do not memorize these as a ritual. The first command exposes desired and observed state, the second connects conditions and events, the third supplies a timeline, and the fourth checks the server's schema. Compare what the controller was asked to do with what it reports doing. Then test the narrowest hypothesis.

Scenario: A rolling update to a StatefulSet gets stuck because the Pod at ordinal 2 is broken and never becomes ready, and because updates proceed strictly one ordinal at a time by default, no lower-ordinal Pod is touched until ordinal 2 succeeds - the rollout is correctly refusing to proceed past a known-bad member rather than plowing ahead. Repeatedly deleting that Pod hoping it "just restarts fixed" wastes time without addressing the actual cause; instead, inspect the current vs. update revision, check whether a partition value is holding the rollout back deliberately, and read that specific Pod's readiness failure reason before doing anything else.

Understand ordered rollout stalls

Default OrderedReady creates and updates members in ordinal order, while StatefulSet rolling updates normally proceed from highest ordinal downward. A partition holds lower ordinals on the old revision for a staged release. Compare currentRevision, updateRevision, Pod labels and readiness. If one ordinal is broken, later progress may stop by design.

Scenario: An updated follower never becomes ready because it waits for a leader at a hostname changed by configuration. Repeated deletion recreates the same bad template. Fix declared configuration and observe membership before proceeding.
Tip: Parallel podManagementPolicy can speed independent startup but cannot be changed into a substitute for application-safe bootstrap. Use it only when ordinal readiness has no prerequisite relationship.

Production reasoning

Ask four questions: Who owns this object? What dependency must become ready next? Which controller reports the blocking condition? What evidence would disprove my current theory? This prevents symptom-driven changes. Record the context, namespace, object generation, image digest and recent rollout before mutation; a recreated Pod may erase the evidence you needed.

Warning: Running is not the same as ready, healthy, durable or correct. Kubernetes status is layered. Confirm the application-level outcome as well as the object state.
Goal: Put this model into practice in the Kubernetes lab statefulset-redis-identity. Open /labs/kubernetes, choose statefulset-redis-identity, predict the failure path before changing anything, then use the simulator's check command to validate the finished state.

Deliberate practice

Before the lab, write the expected object relationship and the first three commands you will run. Afterward, explain why the fix converged and name one tempting change that would only mask the symptom. Repeat using an explicit namespace and a structured output format. This prediction-observation-explanation loop is what turns command familiarity into production judgment.

45-minute investigation

  1. Map (5 min): draw the owner-to-child chain and mark every namespace, selector, identity and dependency involved.
  2. Predict (5 min): write one expected status condition, one likely event and one log or metric signal before opening the lab.
  3. Observe (10 min): collect YAML, describe output and ordered events. Do not mutate state. Record which observation disproves your first theory.
  4. Repair (15 min): make the smallest declarative correction, watch the responsible controller converge, and verify the user-facing path rather than stopping at Running.
  5. Stress (10 min): change one relevant constraint - replica count, label, readiness, resource value or placement rule - predict the outcome, observe it, then restore the known-good declaration.
Tip: Keep a short incident note with symptom, evidence, hypothesis, change, result. Across four sections this produces a reusable runbook instead of a pile of remembered commands.