Kubernetes Jobs CronJobs Guide

💡 Quick Answer: A Job runs a pod to completion (exit 0) and tracks success. A CronJob creates Jobs on a schedule (cron syntax). Use backoffLimit for retry control, parallelism for concurrent execution, ttlSecondsAfterFinished for auto-cleanup, and concurrencyPolicy to control overlapping CronJob runs.

The Problem

Not all workloads are long-running services. Kubernetes needs to handle:

• Database backups on a schedule
• ETL pipelines and data processing
• Report generation
• Email campaigns
• Cleanup and maintenance tasks
• One-off migrations

The Solution

Basic Job

apiVersion: batch/v1
kind: Job
metadata:
  name: db-backup
spec:
  ttlSecondsAfterFinished: 3600   # Auto-delete after 1 hour
  backoffLimit: 3                  # Retry 3 times on failure
  activeDeadlineSeconds: 600       # Timeout after 10 minutes
  template:
    spec:
      containers:
      - name: backup
        image: postgres:16
        command:
        - pg_dump
        - -h
        - postgres-svc
        - -U
        - admin
        - -d
        - production
        - -f
        - /backup/dump.sql
        volumeMounts:
        - name: backup
          mountPath: /backup
      volumes:
      - name: backup
        persistentVolumeClaim:
          claimName: backup-pvc
      restartPolicy: OnFailure

Parallel Job

apiVersion: batch/v1
kind: Job
metadata:
  name: image-processor
spec:
  completions: 10       # Total pods to run successfully
  parallelism: 3        # Run 3 pods at a time
  completionMode: Indexed  # Each pod gets $JOB_COMPLETION_INDEX
  template:
    spec:
      containers:
      - name: processor
        image: myapp/processor:v1
        command:
        - python
        - process.py
        - --shard=$(JOB_COMPLETION_INDEX)
      restartPolicy: Never

CronJob

apiVersion: batch/v1
kind: CronJob
metadata:
  name: nightly-backup
spec:
  schedule: "0 2 * * *"          # 2 AM daily
  timeZone: "Europe/Rome"        # v1.27+ timezone support
  concurrencyPolicy: Forbid      # Don't overlap
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 5
  startingDeadlineSeconds: 600   # Skip if 10min late
  jobTemplate:
    spec:
      ttlSecondsAfterFinished: 86400
      template:
        spec:
          containers:
          - name: backup
            image: backup-tool:v1
            command: ["/backup.sh"]
          restartPolicy: OnFailure

Concurrency Policies

Monitor Jobs

# List jobs
kubectl get jobs
kubectl get cronjobs

# Watch job pods
kubectl get pods -l job-name=db-backup -w

# Check job status
kubectl describe job db-backup

# View logs
kubectl logs job/db-backup

# Manually trigger a CronJob
kubectl create job manual-backup --from=cronjob/nightly-backup

Common Issues

Job pods remain after completion

Set ttlSecondsAfterFinished to auto-cleanup. Without it, completed pods stay until manually deleted or hit successfulJobsHistoryLimit.

CronJob creates duplicate runs

Using Allow concurrency with slow jobs. Switch to Forbid to prevent overlap.

Job never completes — stuck at 0/1

Pod is CrashLooping. Check backoffLimit — once exhausted, the Job fails permanently. Check pod logs for the error.

Best Practices

• Always set ttlSecondsAfterFinished — prevent pod accumulation
• Use Forbid for CronJobs that shouldn’t overlap
• Set activeDeadlineSeconds — prevent runaway jobs
• Use startingDeadlineSeconds on CronJobs — skip stale schedules after downtime
• restartPolicy: Never for debugging — keeps failed pods for log inspection
• restartPolicy: OnFailure for production — automatic retry within the pod

Key Takeaways

• Jobs run pods to completion with retry and parallelism support
• CronJobs create Jobs on a cron schedule with overlap control
• ttlSecondsAfterFinished prevents completed pod accumulation
• concurrencyPolicy: Forbid prevents overlapping scheduled runs
• Use Indexed completion mode for sharded parallel processing
• activeDeadlineSeconds is your timeout safety net