Deploying Jobs

Jobs are saved on the server after Save is pressed in the editor. The visual editor’s Close Editor button will prompt you when trying to close a job with unsaved changes. Closing the visual editor shows the job view:

The job needs to be staged before it can be added to workers. Click Stage job.

The Built-in Worker is now ready and the add button goes green.

Click add; the job runs and the final state reads Deployment Done.

Clicking on Logs in the top right row, shows you that the job started.

Return to the job view (browser back arrow or Alt+Left) and click the red remove button.

This is not a particularly interesting job yet - it runs once, and the output goes to the standard output of the server.

2024-12-17T12:29:19.422Z INFO  lyftdata_worker::scheduler > Worker [Built-In Worker], starting job: [first-job]
{"sample":"data","ts":"2024-12-17T12:29:20.973Z"}

So let’s introduce a problem. Go back to the visual editor, and add a Convert action. This will take existing fields (usually text) and convert them into numbers and so forth. But what if the field does not exist?

Now (and this is important) click on the green Add item button to actually add the conversion.

So Close Action, Save, Close Editor and then Restage Job

Notice that the Issues icon on the top right has gone red. Click on that, and refer back to Logs & Issues to understand how warnings from this job arrive in the panel:

This will also appear in the job logs.