Summary - Job and CronJob

Combine Job and CronJob YAML to write scheduled tasks.

Summary: Job and CronJob

In this group:

A Job runs a task to completion; restartPolicy is Never/OnFailure, and backoffLimit is the retry limit.
activeDeadlineSeconds manages the time, ttlSecondsAfterFinished the cleanup; kubectl wait --for=condition=complete waits.
completions/parallelism give multiple and parallel runs; Indexed mode gives an index per part.
A CronJob creates Jobs on a schedule: cron format, timeZone, jobTemplate.
concurrencyPolicy manages overlap and history limits manage old Jobs; create job --from=cronjob/ triggers by hand.

kubectl create job goc --from=cronjob/gece-yedek
kubectl wait --for=condition=complete job/goc --timeout=300s

Exercise A bounded, self-cleaning Job

Write a Job's spec section (starting with spec:): backoffLimit: 2, activeDeadlineSeconds: 600, ttlSecondsAfterFinished: 3600; restartPolicy: OnFailure in the template; container named rapor, image api:1.0.

Show example solution

spec:
  backoffLimit: 2
  activeDeadlineSeconds: 600
  ttlSecondsAfterFinished: 3600
  template:
    spec:
      restartPolicy: OnFailure
      containers:
        - name: rapor
          image: api:1.0

Test yourself

What happens if you write restartPolicy: Always in a Job pod template?
- The Job behaves like a Deployment
- It is ignored
- The API server rejects it
- The Job never ends
Show answer

The API server rejects it — Only Never and OnFailure are valid in Jobs.

Give the schedule's time zone.

spec:
  schedule: "0 3 * * *"
  ____: "Europe/Istanbul"

Show answer

spec:
  schedule: "0 3 * * *"
  timeZone: "Europe/Istanbul"

Match.

Job

Once, to completion
CronJob

On a schedule
backoffLimit

Retries
suspend

Pause temporarily

Next lesson: What is a PersistentVolume?

We Are Enhancing Your Experience

Cookie Preferences

Strictly Necessary Cookies

Analytical Cookies