Kubernetes: Cron Jobs

There are multiple ways to run a Pod.

a.   Directly run the Pod

b.   Run the Pod using ReplicationSet or ReplicationController.

c.    Run the Pob via Deployment

d.   Run the Pod using a Job

 

You can even run a Pod using Cron Job. CronJobs are used to initialize and start a Pod at specified time.

 

When running a CronJob, a Job will get scheduled, this job will start a Pod.

 

For example, if you want to send email to all your application subscribers on every day morning 5AM, you can create a Cron Job, that spin up a Job at 5 AM, which in turn run a Pod, once the application in the Pod send emails, it terminates successfully. Next day at 5AM, another Job will spin up by cron job and same procedure repeats.

 

cronJobDemo1.yml

apiVersion: batch/v1beta1
kind: CronJob
metadata:
  name: cron-job-to-collect-log-files
  labels:
    app: wakeup-emails
    author: krishna
    serviceType: terminal-app
spec:
  schedule: "*/1 * * * *"
  jobTemplate:
    spec:
      template:
        spec:
          containers:
          - name: collect-log-files
            image: busybox
            command: ["/bin/sh"]
            args: ["-c", "echo 'Collected all the log files'; sleep 10;"]
          restartPolicy: OnFailure

Let’s create a CronJob using above definition file.

$kubectl create -f cronJobDemo1.yml 
cronjob.batch/cron-job-to-collect-log-files created

Let’s confirm that the cronjob, job and a Pod is created.

$kubectl get cronjobs.batch
NAME                            SCHEDULE      SUSPEND   ACTIVE   LAST SCHEDULE   AGE
cron-job-to-collect-log-files   */1 * * * *   False     1        10s             31s
$
$kubectl get pods
NAME                                             READY   STATUS    RESTARTS   AGE
cron-job-to-collect-log-files-1591510980-qbfx9   1/1     Running   0          11s
$
$kubectl get jobs
NAME                                       COMPLETIONS   DURATION   AGE
cron-job-to-collect-log-files-1591510980   0/1           14s        14s

Wait for 10 seconds and query job, pods and you can confirm that the Pod exited successfully and Job is in completed state.

$kubectl get pods
NAME                                             READY   STATUS      RESTARTS   AGE
cron-job-to-collect-log-files-1591510980-qbfx9   0/1     Completed   0          26s
$
$kubectl get jobs
NAME                                       COMPLETIONS   DURATION   AGE
cron-job-to-collect-log-files-1591510980   1/1           17s        29s

Wait for some time (more than a minute) and query jobs and pods, you can confirm new Job is created for every one minute and Pod started to collect log files.

$kubectl get jobs
NAME                                       COMPLETIONS   DURATION   AGE
cron-job-to-collect-log-files-1591510980   1/1           17s        2m56s
cron-job-to-collect-log-files-1591511040   1/1           17s        116s
cron-job-to-collect-log-files-1591511100   1/1           17s        56s
$
$kubectl get pods
NAME                                             READY   STATUS      RESTARTS   AGE
cron-job-to-collect-log-files-1591510980-qbfx9   0/1     Completed   0          3m
cron-job-to-collect-log-files-1591511040-knn2q   0/1     Completed   0          2m
cron-job-to-collect-log-files-1591511100-fvsct   0/1     Completed   0          60s

Let’s stop this cron job by executing below command.

$kubectl delete cronjob.batch cron-job-to-collect-log-files
cronjob.batch "cron-job-to-collect-log-files" deleted

Once you delete a CRON job, all the jobs created by this Cron Job and all the Pods created will get deleted.

$kubectl get cronjob.batch
No resources found in default namespace.
$
$kubectl get jobs
No resources found in default namespace.
$
$kubectl get pods
No resources found in default namespace.

You can refer my below post to know more about CRON expressions.

https://self-learning-java-tutorial.blogspot.com/2014/10/cron-triggers.html

Previous                                                    Next                                                    Home