[Kubernetes] Pod Lifecycle and Probes

Pods follow a defined lifecycle, starting in the Pending phase, moving through Running phase, and then through either the Succeeded or Failed phases.

Pod Lifecycle | Kubernetes

Pod Lifecycle and Probes

Pod Phases

• Pending
  • When a pod is first created, it is in a Pending state.
  • The pod has been accepted by the cluster, but one or more of the containers has not been set up and made ready to run.
  • If a pod is stuck in a pending state, run the kubectl describe pod command, and it will tell you exactly why.
• Running
  • The pod is bound to a node.
  • Once all the containers in a pod start, it goes into a running state.
• Succeeded
  • All containers in the pod have terminated successfully.
• Failed
  • At least one container has terminated in failure.
• Unknown
  • The state of the pod cannot be obtained.

Container States

• Waiting
  • The container is still running the startup operations
• Running
  • The container is executing without issues.
• Terminated
  • The execution has been completed or failed.

Pod Conditions

• PodScheduled
  • When a pod is scheduled on a Node, the PodScheduled condition is set to true.
• ContainersReady
  • When all the containers in the pod are ready, the ContainersReady condition is set to true and finally the pod itself is considered to be Ready.
• Initialized
  • All init containers have started successfully.
• Ready
  • The Ready condition indicates that the applications inside the pod are running and are ready to accept user traffic.

Readiness/Liveness Probes

You can specify the mechanism to check whether a pod is ready (when starting) and healthy (periodically). Probes determine the status of containers.

• Readiness Probe: determines if a pod can start to receive requests
• Liveness Probe: determines if a pod is healthy and running as expected

Probe Actions

• HTTP Test
• TCP Test
• Execute Command

Probe Result

• Success
• Failure
• Unknown

spec:
  containers: 
  - name: my-web
    readinessProbe:  
      httpGet:
        path: /api/ready
        port: 8001
    livenessProbe:  
      httpGet:
        path: /index.html
        port: 8001
      initialDelaySeconds: 15  # wait 15 seconds
      timeoutSeconds: 2  # timeout after 2 seconds 
      periodSeconds: 10  # check every 10 seconds 
      failureThreshold: 3  # allow 3 failures

spec:
  containers:  
  - name: my-app
    readinessProbe:    # or livenessProbe:
      tcpSocket:
        port: 3001

spec:
  containers: 
  - name: my-web
    readinessProbe:    # or livenessProbe:
      exec:
        command: 
          - cat
          - /api/ready