Batch Workloads

Picture a factory. Most machines run continuously, producing goods day and night. But some tasks are different: a machine that only runs once to cut a special piece, or one that runs every morning at 6 AM to prepare raw materials. Kubernetes faced the same need. Not all workloads are long‑running services - some are one‑time tasks or scheduled jobs.

To handle these patterns, Kubernetes introduced Jobs and CronJobs - controllers designed for batch and scheduled workloads.

Jobs – One‑Time Tasks

• Definition: A Job ensures that a pod runs to completion.
• Behavior: If a pod fails, the Job controller creates a new one until the task succeeds.
• Use Cases:
  • Data processing tasks.
  • Database migrations.
  • Report generation.
• Analogy: Jobs are like special orders in a factory - run once, finish, and move on.

CronJobs – Scheduled Tasks

• Definition: A CronJob creates Jobs on a repeating schedule, defined with cron syntax.
• Behavior: Runs tasks at fixed times (e.g., daily, hourly, weekly).
• Use Cases:
  • Nightly backups.
  • Log rotation.
  • Automated report generation.
• Analogy: CronJobs are like alarm clocks for workloads - triggering tasks at precise intervals.

Global Context

• Enterprises: Use Jobs for one‑time tasks like database migrations, and CronJobs for recurring tasks like backups.
• Cloud Providers: Managed Kubernetes services integrate CronJobs with monitoring and alerting systems.
• Community: Jobs and CronJobs are taught worldwide as essential patterns for batch workloads.

Hands‑On Exercise

1. Reflect: How do Jobs guarantee completion, while CronJobs guarantee repetition?

Create a CronJob:

apiVersion: batch/v1
kind: CronJob
metadata:
  name: backup-job
spec:
  schedule: "0 0 * * *"
  jobTemplate:
    spec:
      template:
        spec:
          containers:
          - name: backup
            image: busybox
            command: ["echo", "Running backup..."]
          restartPolicy: OnFailure

kubectl apply -f cronjob.yaml
kubectl get cronjobs

Create a Job:

apiVersion: batch/v1
kind: Job
metadata:
  name: data-job
spec:
  template:
    spec:
      containers:
      - name: data-processor
        image: busybox
        command: ["echo", "Processing data..."]
      restartPolicy: Never

kubectl apply -f job.yaml
kubectl get jobs
kubectl get pods

The Hacker’s Notebook

• Jobs are one‑time workers - ensuring tasks finish successfully.
• CronJobs are timekeepers - triggering tasks on schedule.
• Lesson for engineers: Choose Jobs for completion, CronJobs for repetition.
• Hacker’s mindset: Treat Jobs and CronJobs as automation levers. With them, you can schedule backups, migrations, and reports across the globe.