CKAD — Jobs & CronJobs Lab
Lab 1: Parallel Processing Job
Scenario: Create a Job that processes 6 items with 2 Pods running in parallel.
Step 1: Create the Job
apiVersion: batch/v1
kind: Job
metadata:
name: data-processor
spec:
completions: 6
parallelism: 2
template:
spec:
containers:
- name: processor
image: busybox:1.36
command: ["/bin/sh", "-c"]
args: ["echo Processing batch item && sleep 3 && echo Done"]
restartPolicy: Never
Step 2: Deploy and Watch
kubectl apply -f data-processor.yaml
# Watch the Pods — you should see 2 Pods running at a time
kubectl get pods -l job-name=data-processor -w
# Check completions
kubectl get job data-processor
# Expected: COMPLETIONS 6/6
What’s Happening
- Kubernetes creates 2 Pods simultaneously (
parallelism: 2) - As each Pod completes, a new one replaces it
- This continues until 6 Pods have succeeded (
completions: 6) - The Job is marked
Complete
Lab 2: CronJob with History Limits
Scenario: Create a CronJob that runs every minute, keeps 2 successful and 1 failed Job in history, and uses Forbid concurrency.
Step 1: Create the CronJob
kubectl create cronjob minute-check --image=busybox:1.36 \
--schedule="* * * * *" \
-- /bin/sh -c "echo 'Health check at $(date)'"
Step 2: Configure Settings
kubectl patch cronjob minute-check -p '{
"spec": {
"concurrencyPolicy": "Forbid",
"successfulJobsHistoryLimit": 2,
"failedJobsHistoryLimit": 1
}
}'
Step 3: Verify
# Wait a few minutes and check
kubectl get jobs
# You should see at most 2 successful Jobs
kubectl get cronjob minute-check
# Verify SUSPEND is False and ACTIVE shows recent runs
Cleanup
kubectl delete cronjob minute-check
kubectl delete job data-processor
Practice Questions
Q1: Create a Job named math-job that runs the command expr 3 + 5 using the busybox:1.36 image.
Show Answer
```bash kubectl create job math-job --image=busybox:1.36 -- expr 3 + 5 ``` Check the result: ```bash kubectl logs job/math-job # Expected: 8 ```Q2: A Job has backoffLimit: 4 and restartPolicy: Never. The container keeps failing. How many Pods will be created before the Job gives up?
Show Answer
**5 Pods** will be created — 1 original attempt plus 4 retries (backoffLimit). After the 5th Pod fails, the Job is marked as Failed. With `restartPolicy: Never`, each retry creates a new Pod. With `OnFailure`, the container restarts within the same Pod.Q3: Create a CronJob named log-cleaner that runs every day at 3:00 AM, with concurrencyPolicy: Replace.
Show Answer
```bash kubectl create cronjob log-cleaner --image=busybox:1.36 \ --schedule="0 3 * * *" \ -- /bin/sh -c "echo 'Cleaning old logs'" kubectl patch cronjob log-cleaner -p '{"spec":{"concurrencyPolicy":"Replace"}}' ``` Or as YAML: ```yaml apiVersion: batch/v1 kind: CronJob metadata: name: log-cleaner spec: schedule: "0 3 * * *" concurrencyPolicy: Replace jobTemplate: spec: template: spec: containers: - name: cleaner image: busybox:1.36 command: ["/bin/sh", "-c", "echo 'Cleaning old logs'"] restartPolicy: OnFailure ```Q4: How do you manually trigger a CronJob named weekly-report to run immediately?
Show Answer
```bash kubectl create job weekly-report-manual --from=cronjob/weekly-report ``` This creates a one-time Job from the CronJob's template. It does not affect the CronJob's schedule.Q5: What is the difference between completions and parallelism in a Job spec?
Show Answer
- **`completions`** — The total number of successful Pod completions required for the Job to be considered done. Default is 1. - **`parallelism`** — The maximum number of Pods that can run at the same time. Default is 1. Example: `completions: 10, parallelism: 3` means the Job needs 10 successful Pods total, and will run up to 3 at a time.← Previous: Multi-Container Pods Next: Deployments & Rolling Updates →
