Scheduling a Job with Jenkins

Scheduling a Job with Jenkins

Introduction

Scheduling a job in Jenkins is very useful when it comes to developing a great product. Imagine what great advantage is to transform a time-consuming task into an automated job. It doesn’t need any supervision and as long as there are no errors in the workflow, you shouldn’t have any worries.

In today’s article we’ll make a job that will run every five minutes and has the following output:

Hello, World! Today's: <<date>> Here's the current time: <<time>>

But before we dive into the technical part, let’s take a moment to consider what can be automated in Jenkins. Of course, with a little imagination, you can tackle quite every problem and automate it. Let’s take a very creative example that could be automated with Jenkins if you have a little (more) patience and you’re also eager to deepen your Jenkins skills.

In this example, we’ll talk about Amazon Echo and Alexa. Interesting so far, right? Let’s continue then because Alexa skills could be configured to call jobs that do things. We could have the following steps:

1. Say, “Alexa, lock the doors” (that would be the trigger, meaning Jenkins will listen to that event)
2. Alexa skill activates a Jenkins job.
3. Jenkins job will use a script to call the smart doors locking system API. If the door is already locked do nothing, else try to close the garage door.
4. Alexa waits for the Jenkins job to finish and tells you how it went, i.e. “Door successfully locked” or “The doors could not be locked”.

Of course, that’s not a CI/CD thing but is well within the domain of possibilities for existing tech with Jenkins.

Requirements

• A Jenkins server — for this tutorial, I’ll use the official Docker image to set up a Jenkins server
• An idea of a job that needs to be automated — don’t worry, you can “steal” mine

Creating the Job

Login to your Jenkins application and click on Create a Job. Remember, the type of job doesn’t matter. It can be Freestyle, Maven, etc. All that matters is how we configure the scheduler. For this article, I’ll go with Pipeline. Choose a name for your project and save it.

Now, before we deal with the scheduling part, let’s set up the Pipeline Groovy script which will print the message from the introduction.

Scroll down to Pipeline and add the following script:

The script has three stages: Write Greetings, Write Date, and Write Hour. We use stages in order to better define and structure our script. Also, the stages can be viewed in the Jenkins UI. In case one of them fails, you’ll know which one.

Running the Job

We have our job ready to be triggered! Let’s click on Build Now and check the output.

Started by user Alex Running in Durability level: MAX_SURVIVABILITY Running on Jenkins in /var/jenkins_home/workspace/Scheduled Pipeline Hello, World! Today's: 24-12-2020 Here's the current time: 16:00:38 Finished: SUCCESS

You can also see on the Job’s page the average time for every stage.

Scheduling the Job

There are many ways to trigger a Jenkins job: manually (of course), using a Cron Expression, hooks, watching other projects (an example for this is below), etc. Right now, let’s choose Build periodically and for the Schedule field use the following expression:

H/5 * * * *

The above expression means the job will be executed every 5 minutes. A few (maybe?) useful examples for you:

• 0 0 13 1/1 * ? * – every hour, starting with 13:00;
• 0 0 17 ? * MON-FRI * – every working day, at 5 PM;

If you don’t want to learn this type of formatting in order to schedule your job, you can use a Cron Expression generator like this one.

Let’s see the result. For this, we’ll have to wait. After 10 minutes you should see that 2 jobs were triggered automatically. I let my Jenkins server alone for a few more minutes and here’s the result:

Notice the first job was executed at 9:12 PM, then at 9:17 PM, and the last at 9:22 PM. So the rule worked! We have a job which runs every 5 minutes.

Scheduling a Job by Watching Other Projects

Now that we’ve seen how we can schedule a job periodically, let’s create another one that triggers itself every time the first one was successful. Create another Pipeline project, and add the following code:

We have our pipeline script, let’s schedule it. In the Build Triggers tab, select Build after other projects are built. Type the project which needs to be watched and select the cases when this one should be triggered.

Save it and let’s wait for a little. You should see the new job trigger itself every time the first one succeeds. You can see the details of the first run of this second job that was triggered by the job created before.

Here’s the output:

Started by upstream project "Scheduled Pipeline" build number 7 originally caused by: Started by timer Running in Durability level: MAX_SURVIVABILITY Running on Jenkins in /var/jenkins_home/workspace/The other job worked The other job worked! Finished: SUCCESS

Conclusion

Well, we learned that automation a job with Jenkins might be useful, regardless of the process you want to automate — to build without any input from the users (developers). We saw the great advantage of transforming a time-consuming task into an automated task. The job won’t need supervision and you will have more time for doing tasks that actually require more attention in order for a good quality product to be developed.