Skip to main content

Step 2: Make a Change

Now let's modify the pipeline to see how quickly you can iterate.

Update the Pipeline​

Update hello.yaml:

input:
  generate:
    interval: 1s
    mapping: |
      root.message = "Hello, World!"
      root.timestamp = now()
      root.count = counter()

output:
  stdout: {}

Copy the changed pipeline config under the config: key of hello-job.yaml, the job spec you wrote in Step 1, then redeploy:

expanso-cli job deploy hello-job.yaml

Deploying an existing job name updates that job in place and the control plane re-evaluates it, so you do not stop it first. --force applies the deploy even when the spec is unchanged; to restart a job with the spec it already has, use expanso-cli job rerun hello-world.

Now you see:

{"count":1,"message":"Hello, World!","timestamp":"2024-12-26T10:00:00.123Z"}
{"count":2,"message":"Hello, World!","timestamp":"2024-12-26T10:00:01.456Z"}
{"count":3,"message":"Hello, World!","timestamp":"2024-12-26T10:00:02.789Z"}

What Changed?​

  • now() - A Bloblang function that returns the current timestamp
  • counter() - A function that increments with each message

Try changing the message text or interval and redeploy. Each deploy creates a new job version, so expanso-cli job versions hello-world shows the history and expanso-cli job rollback returns to an earlier one.

What's Next?​

👉 Step 3: Data Transformation