PromptHub
Back to Errors
DevOps Error CrashLoopBackOff Intermediate 12 views

Kubernetes CrashLoopBackOff

A container is repeatedly crashing and restarting in Kubernetes.

Explanation

CrashLoopBackOff occurs when a Kubernetes pod's container is crashing repeatedly and Kubernetes is backing off the restart attempts. The container starts, crashes, Kubernetes waits an increasing amount of time (exponential backoff), then tries again. Common causes include application errors (missing config, database connection failure), misconfigured command/args, missing environment variables, insufficient resources, application bug causing immediate crash, and liveness probe failures. The pod never reaches a running state.

Common Causes

  • Application crashes on startup
  • Missing environment variables or config
  • Database/dependency unreachable
  • Incorrect container command/args
  • Liveness probe failing

Solution

Check the container logs: kubectl logs pod-name. Check previous crashed container logs: kubectl logs pod-name --previous. Describe the pod for events: kubectl describe pod pod-name. Verify the container command/args are correct. Check that all required environment variables and config maps are present. Ensure database and other dependencies are accessible. Verify the container image works locally. Check resource limits - the container may be OOMKilling. Review liveness/readiness probe configuration. Check for application-level errors in the logs.

Code Example 

# Check pod status
kubectl get pods
# NAME    READY   STATUS             RESTARTS   AGE
# myapp   0/1     CrashLoopBackOff   5          2m

# Check current logs
kubectl logs myapp-pod-xyz

# Check logs from crashed container
kubectl logs myapp-pod-xyz --previous

# Describe pod for events
kubectl describe pod myapp-pod-xyz
# Look at Events section for errors

# Common fixes:
# 1. Missing config
kubectl get configmap myapp-config

# 2. Database connection - check service
kubectl get svc postgres
kubectl exec -it myapp-pod-xyz -- env | grep DATABASE

# 3. Check command/args
apiVersion: v1
kind: Pod
spec:
  containers:
  - name: myapp
    image: myapp:latest
    command: ["node"]
    args: ["server.js"]  # verify this file exists!

# 4. Check resource limits
kubectl top pod myapp-pod-xyz

# 5. Test image locally
docker run --rm myapp:latest

Error Information

Language

bash

Framework

kubernetes

Difficulty

Intermediate

Views

12

Related Errors

DevOps

Docker Permission Denied

The current user doesn't have permission to run Docker commands.
General

Git Merge Conflict

Git cannot automatically merge changes from two branches because they modified the same lines.
DevOps

Container Name Already in Use

A Docker container with the specified name already exists.
DevOps

Ansible Unreachable Host

Ansible cannot connect to the target host via SSH or WinRM.