Skip to main content

Lab 7.3 - Jobs and CronJobs

Lab Objectives​

By the end of this lab, you will be able to:

  • Create and run Jobs for one-time tasks.
  • Understand different types of Jobs (complete, parallel).
  • Create CronJobs for scheduled tasks.
  • Manage Job history.
  • Monitor the execution of Jobs and CronJobs.

Estimated Duration​

45-60 minutes

Prerequisites​

  • kubectl installed and configured.
  • Functional local Kubernetes cluster (minikube or kind).
  • Knowledge of Jobs and CronJobs (Chapter 7.3).

Part 1: Creating a Simple Job​

A Job creates one or more Pods and ensures that a specified number of them successfully complete.

Step 1.1: Create a Simple Job​

Create a file job-simple.yaml:

# job-simple.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: job-simple
spec:
  template:
    spec:
      containers:
      - name: task
        image: busybox:1.35
        command: ["/bin/sh", "-c"]
        args: ["echo 'Task completed successfully!' && date"]
      restartPolicy: Never # Jobs use Never or OnFailure

Explanation:

  • restartPolicy: Never: The Pod will not be automatically restarted if it fails. For Jobs, use Never or OnFailure.
  • The Job creates a Pod that runs the command and completes.

Apply the Job:

kubectl apply -f job-simple.yaml

Step 1.2: Verify Execution​

Check the status of the Job:

kubectl get jobs

Check the created Pods:

kubectl get pods -l job-name=job-simple

Once the Pod is completed, check the logs:

kubectl logs -l job-name=job-simple

You should see the message "Task completed successfully!" and the date.

Step 1.3: Verify Complete Status​

Check the details of the Job:

kubectl describe job job-simple

You should see that the Job is Complete with 1/1 completed.

Part 2: Job with Multiple Completions​

A Job can be configured to run multiple times (in parallel or sequentially).

Step 2.1: Create a Job with Completions​

Create a file job-multiple.yaml:

# job-multiple.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: job-multiple
spec:
  completions: 5 # The Job must run successfully 5 times
  parallelism: 2 # Run 2 Pods in parallel at a time
  template:
    spec:
      containers:
      - name: task
        image: busybox:1.35
        command: ["/bin/sh", "-c"]
        args: ["echo 'Task $(date +%s)' && sleep 2"]
      restartPolicy: Never

Explanation:

  • completions: 5: The Job must complete 5 successful runs.
  • parallelism: 2: Maximum 2 Pods run in parallel at a time.

Apply the Job:

kubectl apply -f job-multiple.yaml

Step 2.2: Observe Execution​

Monitor the Pods in real-time:

kubectl get pods -l job-name=job-multiple -w

You should see 2 Pods running in parallel, then others being created until 5 completions are reached.

Check the status of the Job:

kubectl get job job-multiple

Once completed, you should see 5/5 completions.

Part 3: Job with Backoff Limit​

The backoffLimit defines the number of retry attempts in case of failure.

Step 3.1: Create a Failing Job​

Create a file job-failure.yaml:

# job-failure.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: job-failure
spec:
  backoffLimit: 3 # Maximum 3 retry attempts in case of failure
  template:
    spec:
      containers:
      - name: task
        image: busybox:1.35
        command: ["/bin/sh", "-c"]
        args: ["exit 1"] # Intentionally fails
      restartPolicy: Never

Apply the Job:

kubectl apply -f job-failure.yaml

Step 3.2: Observe the Attempts​

Monitor the Pods:

kubectl get pods -l job-name=job-failure -w

You should see multiple Pods being created (up to 3 attempts) before the Job is marked as Failed.

Check the status of the Job:

kubectl get job job-failure
kubectl describe job job-failure

Part 4: Creating a CronJob​

A CronJob creates Jobs recurrently according to a schedule (cron format).

Step 4.1: Create a CronJob​

Create a file cronjob-example.yaml:

# cronjob-example.yaml
apiVersion: batch/v1
kind: CronJob
metadata:
  name: cronjob-example
spec:
  schedule: "*/1 * * * *" # Every minute (for testing)
  jobTemplate:
    spec:
      template:
        spec:
          containers:
          - name: task
            image: busybox:1.35
            command: ["/bin/sh", "-c"]
            args: ["echo 'CronJob execution - $(date)'"]
          restartPolicy: OnFailure
  successfulJobsHistoryLimit: 3 # Keep 3 successful Jobs in history
  failedJobsHistoryLimit: 1 # Keep 1 failed Job in history

Explanation:

  • schedule: "*/1 * * * *": Cron format (minute hour day month day-of-week). Here, every minute.
  • successfulJobsHistoryLimit: 3: Keeps the last 3 successful Jobs.
  • failedJobsHistoryLimit: 1: Keeps the last failed Job.

Apply the CronJob:

kubectl apply -f cronjob-example.yaml

Step 4.2: Verify the CronJob​

Check the status of the CronJob:

kubectl get cronjob cronjob-example
# or
kubectl get cj cronjob-example

Wait a minute, then check that Jobs are being created:

kubectl get jobs

You should see Jobs named cronjob-example-<timestamp>.

Check the Pods created by the Jobs:

kubectl get pods

Check the logs of one of the Jobs:

JOB_NAME=$(kubectl get jobs -o jsonpath='{.items[0].metadata.name}')
kubectl logs -l job-name=$JOB_NAME

Part 5: Suspending a CronJob​

You can temporarily suspend a CronJob without deleting it.

Step 5.1: Suspend the CronJob​

Suspend the CronJob:

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

Check that the CronJob is suspended:

kubectl get cronjob cronjob-example

You should see True in the SUSPEND column.

Step 5.2: Resume the CronJob​

Resume the CronJob:

kubectl patch cronjob cronjob-example -p '{"spec":{"suspend":false}}'

Part 6: Examples of Cron Schedules​

Here are some common cron schedule examples:

schedule: "0 0 * * *"        # Every day at midnight
schedule: "0 */6 * * *"     # Every 6 hours
schedule: "0 9 * * 1"       # Every Monday at 9 AM
schedule: "0 0 1 * *"       # The first day of every month at midnight
schedule: "*/30 * * * *"    # Every 30 minutes

Part 7: Cleanup​

Delete the created resources:

# Delete the CronJobs (also deletes associated Jobs)
kubectl delete cronjob cronjob-example

# Delete the Jobs
kubectl delete job job-simple job-multiple job-failure

# Delete remaining Pods (optional, normally deleted automatically)
kubectl delete pods -l job-name

Lab Summary​

In this lab, you explored Jobs and CronJobs. You learned how to create Jobs for one-time tasks, configure parallel Jobs, manage failures with backoffLimit, and create CronJobs for recurring scheduled tasks.

Next Steps​

The final lab of this module will show you how to use Init Containers and Sidecars for advanced patterns.

Lab 7.4: Init Containers and Sidecars

Lab created on: December 2024