Kubernetes CronJobs

The time-based scheduler for Kubernetes batch workloads. A CronJob wraps a Job with a cron expression, creating a new Job (and therefore new Pods) on every schedule tick. Synthesized from CKA Day 12 — DaemonSet, Job & CronJob Explained.

What is a CronJob?

A CronJob is a Kubernetes controller that creates Jobs on a recurring schedule. It is the cluster-native equivalent of crontab — but with the full power of Kubernetes resource management, fault tolerance, and observability.

Every time the schedule triggers, the CronJob controller spawns a new Job object. That Job creates Pods, waits for them to complete, and then reports status. Because each scheduled run is an independent Job, a failed run does not block the next scheduled execution.

Key Insight: CronJob → Job → Pod. The CronJob never creates Pods directly. It is a meta-controller that manages Job lifecycles over time.

Why Use a CronJob?

Requirement	Linux crontab	Kubernetes CronJob
Run on a schedule	✅	✅
Cluster-aware resource limits	❌	✅
Automatic retry on failure	❌	✅ (via Job backoffLimit)
Run history and logs	❌ (manual)	✅ (kubectl get jobs)
Concurrency control	❌	✅ (concurrencyPolicy)
Suspend without deleting	❌	✅ (suspend: true)
History cleanup	❌	✅ (successfulJobHistoryLimit)

Canonical Use Cases

• Scheduled backups: Database dumps, etcd snapshots, volume snapshots
• Periodic cleanup: Prune old logs, delete temporary files, expire cache entries
• Recurring reports: Daily sales reports, weekly analytics exports, monthly audits
• Health checks: Periodic liveness tests, certificate expiration checks, compliance scans
• Data synchronization: Nightly ETL syncs, batch imports from external APIs

YAML Structure

apiVersion: batch/v1
kind: CronJob
metadata:
  name: db-backup
spec:
  schedule: "0 2 * * *"  # Daily at 2:00 AM
  startingDeadlineSeconds: 3600
  concurrencyPolicy: Forbid
  successfulJobHistoryLimit: 3
  failedJobHistoryLimit: 1
  suspend: false
  jobTemplate:
    spec:
      backoffLimit: 3
      activeDeadlineSeconds: 1800
      template:
        spec:
          restartPolicy: OnFailure
          containers:
          - name: backup
            image: postgres:15
            command:
            - pg_dump
            - -h
            - db-service
            - -U
            - admin
            - -f
            - /backups/daily.sql
            volumeMounts:
            - name: backup-vol
              mountPath: /backups
          volumes:
          - name: backup-vol
            emptyDir: {}

Key fields:

Field	Default	Description
schedule	Required	Cron expression (standard 5-field format)
startingDeadlineSeconds	unset	Max seconds after scheduled time to still start the Job
concurrencyPolicy	Allow	Allow (overlap), Forbid (skip if running), Replace (kill old, start new)
successfulJobHistoryLimit	3	Number of completed Job objects to retain
failedJobHistoryLimit	1	Number of failed Job objects to retain
suspend	false	Pause the schedule without deleting the CronJob
jobTemplate	Required	The Job spec template for each scheduled run

Cron Expression Format

Kubernetes uses the standard 5-field cron format:

┌───────────── minute (0 - 59)
│ ┌───────────── hour (0 - 23)
│ │ ┌───────────── day of month (1 - 31)
│ │ │ ┌───────────── month (1 - 12)
│ │ │ │ ┌───────────── day of week (0 - 6, Sunday = 0)
│ │ │ │ │
* * * * *

Common patterns:

Expression	Meaning
*/5 * * * *	Every 5 minutes
0 * * * *	Every hour at minute 0
0 2 * * *	Daily at 2:00 AM
0 0 * * 0	Every Sunday at midnight
0 0 1 * *	First day of every month at midnight

Concurrency Policies

Policy	Behavior	When to Use
Allow	Multiple Jobs from the same CronJob can run simultaneously	Independent, idempotent tasks
Forbid	If a Job is still running, the next scheduled start is skipped	Tasks that must not overlap
Replace	If a Job is still running, it is terminated and a new one starts	Always need the latest data

Managing History and Cleanup

By default, Kubernetes keeps 3 successful Jobs and 1 failed Job. Older Jobs are automatically deleted. You can adjust these limits or use a ttlSecondsAfterFinished field inside the jobTemplate to delete individual Jobs (and their Pods) after a set time.

Suspending a CronJob

kubectl patch cronjob db-backup -p '{"spec":{"suspend":true}}'

This stops the schedule without deleting the CronJob definition — useful for maintenance windows or incident response.

Essential Commands

Command	Purpose
kubectl get cronjobs	List CronJobs
kubectl get jobs	See Jobs created by the CronJob
kubectl describe cronjob <name>	Schedule, history, and last run status
kubectl logs job/<job-name>	Read logs from a specific run
kubectl delete cronjob <name>	Delete CronJob and stop scheduling
kubectl create cronjob ... --dry-run=client -o yaml	Generate a manifest quickly

CKA Exam Relevance

• Workloads & Scheduling (~15%): Create a CronJob with a given schedule. Be precise with the cron expression.
• Troubleshooting (~30%): If a CronJob missed a run, check startingDeadlineSeconds (was the controller down too long?), concurrencyPolicy (was it Forbid and a previous Job still running?), and suspend status.
• Speed pattern:

  kubectl create cronjob backup --image=busybox --schedule="0 2 * * *" \
    --dry-run=client -o yaml > cronjob.yaml
  # Edit jobTemplate to add volumes, env vars, and restartPolicy: OnFailure

Sources

• CKA Day 12 — DaemonSet, Job & CronJob Explained

Related Pages

• Kubernetes Jobs — the unit of execution a CronJob manages
• Pod Fundamentals — the underlying compute unit
• Deployment, ReplicaSet & Replication Controller — long-running controllers
• Kubernetes Namespaces — CronJobs are namespace-scoped
• Kubernetes Labels and Selectors — label-based filtering for CronJob targets
• CKA Certification — exam overview
• CKA Study Roadmap — 40-day plan
• Tech Tutorials with Piyush — course source

---

Tags: kubernetes cronjob batch scheduling cka devops backup cleanup cron