How to Fix CrashLoopBackOff Pod Status in Kubernetes

🔍 Quick Diagnostics

🚀 How to Fix the Error Step-by-Step

Step 1: Inspect the failure reason with kubectl describe

The primary command for gathering lifecycle details about the failing Pod is kubectl describe:

# Fetch detailed status of the target Pod
kubectl describe pod <pod-name> -n <namespace>

In the Containers > State and Last State sections, pay close attention to the following indicators:

• Exit Code 1 or 127: Application failure (unhandled runtime exception, missing file, or broken CMD path).
• Exit Code 137 (OOMKilled): The container exceeded its strict memory limit (limits.memory) configured in the YAML manifest.
• Liveness probe failed: The configured health check failed to respond within the specified timeout.

Step 2: Extract logs from the crashed container

To inspect standard output (stdout/stderr) right before the container crashed on its previous attempt, add the --previous flag:

# View logs from the previous container instance
kubectl logs <pod-name> --previous -n <namespace>

# If the Pod has multiple containers, specify the container name:
kubectl logs <pod-name> -c <container-name> --previous -n <namespace>

Step 3: Adjust container memory limits (OOMKilled)

If the container was killed due to memory exhaustion (exit code 137), update your deployment manifest to allocate additional memory resources:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-service
spec:
  template:
    spec:
      containers:
      - name: my-app
        image: my-image:latest
        resources:
          requests:
            memory: "256Mi"
            cpu: "250m"
          limits:
            memory: "512Mi" # Increase limit if OOMKilled occurs
            cpu: "500m"

Apply the changes:

kubectl apply -f deployment.yaml

Step 4: Adjust or delay Liveness and Readiness Probes

If your application takes a long time to boot (such as performing database migrations on startup), the livenessProbe might kill the Pod prematurely. Increase initialDelaySeconds:

livenessProbe:
  httpGet:
    path: /healthz
    port: 8080
  initialDelaySeconds: 30 # Give the application time to complete initialization
  periodSeconds: 10
  failureThreshold: 3