Container Probes

Probe types

• Order: startup first → then readiness + liveness concurrently

All three probes with all config fields

spec:
  containers:
  - name: app
    image: myapp:1.0
    livenessProbe:
      httpGet:
        path: /healthz
        port: 8080
      initialDelaySeconds: 10   # wait before first probe
      periodSeconds: 10         # how often to probe
      timeoutSeconds: 1         # probe timeout
      failureThreshold: 3       # failures before action
      successThreshold: 1       # successes to mark healthy (liveness: always 1)
    readinessProbe:
      tcpSocket:
        port: 5432
      initialDelaySeconds: 5
      periodSeconds: 5
      failureThreshold: 3
      successThreshold: 1       # readiness: can be >1
    startupProbe:
      exec:
        command:
        - cat
        - /tmp/ready             # exit 0 = success
      failureThreshold: 30      # 30 × 10s = 5 min for slow startup
      periodSeconds: 10

Verification methods

gRPC probe

livenessProbe:
  grpc:
    port: 9090
    service: ""            # optional
  initialDelaySeconds: 10
  periodSeconds: 10

httpGet with HTTPS

livenessProbe:
  httpGet:
    path: /healthz
    port: 8443
    scheme: HTTPS
    httpHeaders:
    - name: Custom-Header
      value: Awesome

Timing defaults

kubectl

kubectl get po <n>        # READY 0/1 if readiness failing
kubectl describe po <n>   # probe config + events

Failures

Exam gotchas

• readinessProbe failing ≠ restart: Pod stays Running but removed from Service endpoints; auto-rejoins when fixed
• Liveness too aggressive on slow apps → CrashLoop: use startupProbe
• httpGet: correct path and port (named port preferred)