Skip to content

Parameterization

In the previous section, we saw how to set up a basic webhook event-source and sensor. The trigger template had parameters set in the sensor object, and the workflow was able to print the event payload. In this tutorial, we will dig deeper into different types of parameterization, how to extract particular key-value from event payload and how to use default values if certain key is not available within event payload.

Trigger Resource Parameterization

If you take a closer look at the Sensor object, you will notice it contains a list of triggers. Each Trigger contains the template that defines the context of the trigger and actual resource that we expect the sensor to execute. In the previous section, the resource within the trigger template was an Argo workflow.

This subsection deals with how to parameterize the resource within trigger template with the event payload.

Prerequisites

Make sure to have the basic webhook event-source and sensor set up. Follow the introduction tutorial if haven't done already.

Webhook Event Payload

Webhook event-source consumes events through HTTP requests and transforms them into CloudEvents. The structure of the event the Webhook sensor receives from the event-source over the eventbus looks like following,

    {
        "context": {
          "type": "type_of_event_source",
          "specversion": "cloud_events_version",
          "source": "name_of_the_event_source",
          "id": "unique_event_id",
          "time": "event_time",
          "datacontenttype": "type_of_data",
          "subject": "name_of_the_configuration_within_event_source"
        },
        "data": {
          "header": {},
          "body": {},
        }
    }
  1. Context: This is the CloudEvent context and it is populated by the event-source regardless of type of HTTP request.

  2. Data: Data contains following fields.

  3. Header: The header within event data contains the headers in the HTTP request that was dispatched to the event-source. The event-source extracts the headers from the request and put it in the header within event data.

  4. Body: This is the request payload from the HTTP request.

Event Context

Now that we have an understanding of the structure of the event the webhook sensor receives from the event-source over the eventbus, lets see how we can use the event context to parameterize the Argo workflow.

  1. Update the Webhook Sensor and add the contextKey for the parameter at index 0.

    kubectl -n argo-events apply -f https://raw.githubusercontent.com/argoproj/argo-events/stable/examples/tutorials/02-parameterization/sensor-01.yaml
    
  2. Send a HTTP request to the event-source pod.

    curl -d '{"message":"this is my first webhook"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example
    
  3. Inspect the output of the Argo workflow that was created.

    argo logs name_of_the_workflow
    

You will see the following output,

    _________
    < webhook >
    ---------
       \
        \
         \
                       ##        .
                 ## ## ##       ==
              ## ## ## ##      ===
          /""""""""""""""""___/ ===
     ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
          \______ o          __/
           \    \        __/
             \____\______/

We have successfully extracted the type key within the event context and parameterized the workflow to print the value of the type.

Event Data

Now, it is time to use the event data and parameterize the Argo workflow trigger. We will extract the message from request payload and get the Argo workflow to print the message.

  1. Update the Webhook Sensor and add the dataKey in the parameter at index 0.

    kubectl -n argo-events apply -f https://raw.githubusercontent.com/argoproj/argo-events/stable/examples/tutorials/02-parameterization/sensor-02.yaml
    
  2. Send a HTTP request to the event-source pod.

    curl -d '{"message":"this is my first webhook"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example
    
  3. Inspect the output of the Argo workflow that was created.

    argo logs name_of_the_workflow
    

You will see the following output,

     __________________________
    < this is my first webhook >
     --------------------------
        \
         \
          \
                        ##        .
                  ## ## ##       ==
               ## ## ## ##      ===
           /""""""""""""""""___/ ===
      ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
           \______ o          __/
            \    \        __/
              \____\______/

Yay!! The Argo workflow printed the message. You can add however many number of parameters to update the trigger resource on the fly.

Note: If you define both the contextKey and dataKey within a parameter, then the dataKey takes the precedence.

Default Values

Each parameter comes with an option to configure the default value. This is specially important when the key you defined in the parameter doesn't exist in the event.

  1. Update the Webhook Sensor and add the value for the parameter at index 0. We will also update the dataKey to an unknown event key.

    kubectl -n argo-events apply -f https://raw.githubusercontent.com/argoproj/argo-events/stable/examples/tutorials/02-parameterization/sensor-03.yaml
    
  2. Send a HTTP request to the event-source pod.

    curl -d '{"message":"this is my first webhook"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example
    
  3. Inspect the output of the Argo workflow that was created.

    argo logs name_of_the_workflow
    

You will see the following output,

    _______________________
    < wow! a default value. >
    -----------------------
       \
        \
         \
                       ##        .
                 ## ## ##       ==
              ## ## ## ##      ===
          /""""""""""""""""___/ ===
     ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
          \______ o          __/
           \    \        __/
             \____\______/


Sprig Templates

The sprig template exposed through contextTemplate and dataTemplate lets you alter the event context and event data before it gets applied to the trigger via parameters.

Take a look at the example defined here, it contains the parameters as follows,

    parameters:
    # Retrieve the 'message' key from the payload
    - src:
        dependencyName: test-dep
        dataTemplate: "{{ .Input.body.message | title }}"
      dest: spec.arguments.parameters.0.value
    # Title case the context subject
    - src:
        dependencyName: test-dep
        contextTemplate: "{{ .Input.subject | title }}"
      dest: spec.arguments.parameters.1.value
    # Retrieve the 'name' key from the payload, remove all whitespace and lowercase it.
    - src:
        dependencyName: test-dep
        dataTemplate: "{{ .Input.body.name | nospace | lower }}-"
      dest: metadata.generateName
      operation: append

Consider the event the sensor received has format like,

    {
        "context": {
          "type": "type_of_event_source",
          "specversion": "cloud_events_version",
          "source": "name_of_the_event_source",
          "id": "unique_event_id",
          "time": "event_time",
          "datacontenttype": "type_of_data",
          "subject": "name_of_the_configuration_within_event_source"
        },
        "data": {
          "body": {
            "name": "foo bar",
            "message": "hello there!!"
          },
        }
    }

The parameters are transformed as,

  1. The first parameter extracts the body.message from event data and applies title filter which basically capitalizes the first letter and replaces the spec.arguments.parameters.0.value.

  2. The second parameter extracts the subject from the event context and again applies title filter and replaces the spec.arguments.parameters.1.value.

  3. The third parameter extracts the body.name from the event data, applies nospace filter which removes all white spaces and then lower filter which lowercases the text and appends it to metadata.generateName.

Send a curl request to event-source as follows,

    curl -d '{"name":"foo bar", "message": "hello there!!"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example

and you will see an Argo workflow being sprung with name like webhook-foobar-xxxxx.

Check the output of the workflow, it should print something like,

     ____________________________
    < Hello There!! from Example >
     ----------------------------
        \
         \
          \
                        ##        .
                  ## ## ##       ==
               ## ## ## ##      ===
           /""""""""""""""""___/ ===
      ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
           \______ o          __/
            \    \        __/
              \____\______/


Operations

Sometimes you need the ability to append or prepend a parameter value to an existing value in trigger resource. This is where the operation field within a parameter comes handy.

  1. Update the Webhook Sensor and add the operation in the parameter at index 0. We will prepend the message to an existing value.

    kubectl -n argo-events apply -f https://raw.githubusercontent.com/argoproj/argo-events/stable/examples/tutorials/02-parameterization/sensor-04.yaml
    
  2. Send a HTTP request to the event-source.

    curl -d '{"message":"hey!!"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example
    
  3. Inspect the output of the Argo workflow that was created.

    argo logs name_of_the_workflow
    

You will see the following output,

     __________________
    < hey!!hello world >
     ------------------
        \
         \
          \
                        ##        .
                  ## ## ##       ==
               ## ## ## ##      ===
           /""""""""""""""""___/ ===
      ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
           \______ o          __/
            \    \        __/
              \____\______/

Trigger Template Parameterization

The parameterization you saw above deals with the trigger resource, but sometimes you need to parameterize the trigger template itself. This comes handy when you have the trigger resource stored on some external source like S3, Git, etc. and you need to replace the url of the source on the fly in trigger template.

Imagine a scenario where you want to parameterize the parameters of trigger to parameterize the trigger resource. What?...

The sensor you have been using in this tutorial has one parameter defined in the trigger resource under k8s. We will parameterize that parameter by applying a parameter at the trigger template level.

  1. Update the Webhook Sensor and add parameters at trigger level.

    kubectl -n argo-events apply -f https://raw.githubusercontent.com/argoproj/argo-events/stable/examples/tutorials/02-parameterization/sensor-05.yaml
    
  2. Send a HTTP request to the event-source.

    curl -d '{"dependencyName":"test-dep", "dataKey": "body.message", "dest": "spec.arguments.parameters.0.value", "message": "amazing!!"}' -H "Content-Type: application/json" -X POST http://localhost:12000/example
    
  3. Inspect the output of the Argo workflow that was created.

    argo logs name_of_the_workflow
    

You will see the following output,

    ___________
    < amazing!! >
    -----------
    \
     \
      \
                    ##        .
              ## ## ##       ==
           ## ## ## ##      ===
       /""""""""""""""""___/ ===
    ~~~ {~~ ~~~~ ~~~ ~~~~ ~~ ~ /  ===- ~~~
       \______ o          __/
        \    \        __/
          \____\______/

Great!! You have now learned how to apply parameters at trigger resource and template level. Keep in mind that you can apply default values and operations like prepend and append for trigger template parameters as well.