Secrets Without Security Theater

A Kubernetes Secret looks and feels different from a ConfigMap, but under the hood it is only base64-encoded - not encrypted - inside the API object itself, and base64 is a reversible encoding, not a cipher; anyone who can read the Secret object can trivially decode its contents. Real protection has to come from several layers working together: RBAC controlling exactly who and what can get or list Secrets, encryption at rest configured on etcd so the raw stored bytes are not plaintext on disk, narrowly-scoped ServiceAccounts so a compromised Pod cannot read secrets it does not need, and, for genuinely sensitive credentials, an external secrets manager that Kubernetes only references rather than storing directly. When an application library supports it, prefer mounting Secrets as files rather than environment variables, since a well-behaved library can reread an updated file, while an environment variable - as covered above for ConfigMaps - is frozen at process start. And two easy but important habits: never let a Secret's decoded value end up in application logs or a shell history, and never commit decoded secret material to Git, even "temporarily."

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 database password is rotated and the new value correctly lands inside the Kubernetes Secret object within seconds - but every already-running Pod loaded the old password into an environment variable when it started, days or weeks ago, and keeps using that stale value until it happens to restart. A rotation plan has to explicitly cover propagation (getting the new value in front of the application), reload (getting the application to actually pick it up), and an overlap window (keeping the old credential valid briefly so in-flight connections do not break) - simply updating the Secret object is only step one of three.

Design the complete secret lifecycle

Creation, storage, distribution, use, rotation and revocation all matter. Enable API encryption at rest, restrict get/list/watch, prefer short-lived external credentials when available, and prevent secret values from entering logs, command histories or rendered diffs. Mounted secret files can update; environment copies cannot. Applications may still hold old credentials in pools, so rotations often need overlap between old and new values.

Scenario: A database password rotates instantly and every existing connection works until reconnect. New connections fail because Pods never reloaded the secret. Monitor both credential versions and force a controlled reconnect before revocation.
Warning: Anyone allowed to create a Pod that mounts a Secret may effectively read it even without direct Secret get permission. Authorization reviews must include workload creation paths.

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 cert-manager-issuer-secret. Open /labs/kubernetes, choose cert-manager-issuer-secret, 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.