Intermediate13 min read

Jobs & CronJobs

Run batch workloads and scheduled tasks in Kubernetes — Jobs for one-off tasks, CronJobs for recurring schedules, and parallel processing patterns.

Kubernetes Jobs

Jobs run pods to completion — unlike Deployments which keep pods running indefinitely. A Job creates one or more pods and tracks successful completions. Use Jobs for batch processing, database migrations, and one-off tasks.

Jobs retry failed pods up to backoffLimit times. Set completions and parallelism for work queue patterns where multiple pods process items concurrently.

restartPolicy must be Never or OnFailure for Jobs
completions: 5 with parallelism: 2 runs 5 pods, 2 at a time
Jobs are not restarted when they complete — create new ones

apiVersion: batch/v1
kind: Job
metadata:
  name: data-migration
spec:
  backoffLimit: 3
  template:
    spec:
      restartPolicy: Never
      containers:
        - name: migrate
          image: myapp:1.0.0
          command: ["npm", "run", "migrate"]

CronJobs

CronJobs create Jobs on a schedule using cron syntax. They manage Job history — successfulJobsHistoryLimit and failedJobsHistoryLimit control retention. Use CronJobs for backups, report generation, cache warming, and cleanup tasks.

CronJob schedules use standard cron format: "0 2 * * *" runs daily at 2 AM UTC. Specify timezone with timeZone field (Kubernetes 1.27+).

apiVersion: batch/v1
kind: CronJob
metadata:
  name: nightly-backup
spec:
  schedule: "0 2 * * *"
  timeZone: "America/New_York"
  concurrencyPolicy: Forbid
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 1
  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: OnFailure
          containers:
            - name: backup
              image: backup-tool:latest
              command: ["./backup.sh"]

Parallel and Indexed Jobs

Parallel Jobs run multiple pods simultaneously with completions tracking total successes. Indexed Jobs assign each pod an index (0, 1, 2...) useful for partitioned work — processing data shards or running distributed tests.

Set completionMode: Indexed for indexed jobs. Pods access their index via JOB_COMPLETION_INDEX environment variable.

spec:
  completions: 10
  parallelism: 3
  completionMode: Indexed
  template:
    spec:
      containers:
        - name: worker
          image: worker:latest
          env:
            - name: SHARD_INDEX
              valueFrom:
                fieldRef:
                  fieldPath: metadata.annotations['batch.kubernetes.io/job-completion-index']

Job Patterns and Best Practices

Use init containers in Job pods for setup tasks before the main container runs. Set activeDeadlineSeconds to kill jobs that run too long. Use ttlSecondsAfterFinished to auto-clean completed jobs.

For long-running batch workloads, consider Argo Workflows or Tekton Pipelines which provide DAG-based orchestration, artifact passing, and retry logic beyond basic Jobs.

spec:
  activeDeadlineSeconds: 3600  # Kill after 1 hour
  ttlSecondsAfterFinished: 86400  # Clean up after 24 hours
  template:
    spec:
      initContainers:
        - name: setup
          image: setup:latest
          command: ["./prepare-data.sh"]

Monitoring Jobs

Monitor job success and failure rates with Prometheus metrics from kube-state-metrics. Alert on failed CronJobs and jobs exceeding deadline. Check job logs with kubectl logs job/job-name.

For CronJobs, a missed schedule (startingDeadlineSeconds exceeded) is logged but not retried. Monitor schedule adherence and job duration trends.