Kogito Serverless Workflow - Knative custom function

Description

This example contains a simple workflow service that illustrates how to use Serverless Workflow custom function to invoke Knative services.

This example consists of a workflow and a remote service that both will be deployed to Knative.

The service is described using JSON format as defined in the CNCF Serverless Workflow specification.

Installing and Running

Prerequisites

You will need:

• Java 17+ installed
• Environment variable JAVA_HOME set accordingly
• Maven 3.8.6++ installed When using native image compilation, you will also need:
• GraalVm 22.3.0+ installed
• Environment variable GRAALVM_HOME set accordingly
• Note that GraalVM native image compilation typically requires other packages (glibc-devel, zlib-devel and gcc) to be installed too. You also need ‘native-image’ installed in GraalVM (using ‘gu install native-image’). Please refer to GraalVM installation documentation for more details.
• minikube (https://minikube.sigs.k8s.io/docs/start/)
• Knative quickstarts

Minikube settings

Execute this command to build the images in your Minikube internal registry.

eval $(minikube -p minikube docker-env --profile knative)

To expose the service URLs in your local environment you must start the minikube tunnel.

# you will be asked for your admin password
minikube tunnel --profile knative

Creating the namespace

# The namespace name is very important to ensure all the services that compose the example can interact.
kubectl create ns custom-functions

Deploy the custom-function-knative-service project to Knative

From the custom-function-knative-service directory run:

mvn clean package -Dcontainer

kubectl apply -f target/kubernetes/knative.yml

Check if the service was successfully deployed:

kn service list -n custom-functions

You should see something similar to (URL should change):

NAME                              URL                                                                             LATEST                                  AGE   CONDITIONS   READY   REASON
custom-function-knative-service   http://custom-function-knative-service.custom-functions.10.107.55.54.sslip.io   custom-function-knative-service-00001   53s   3 OK / 3     True     

Deploy the workflow project to Knative

From the workflow directory run:

mvn clean package -Dcontainer

kubectl apply -f target/kubernetes/knative.yml

Check if the service was successfully deployed:

kn service list -n custom-functions

You should see both services deployed similar to (URLs should change):

NAME                              URL                                                                             LATEST                                  AGE     CONDITIONS   READY   REASON
custom-function-knative-service   http://custom-function-knative-service.custom-functions.10.107.55.54.sslip.io   custom-function-knative-service-00001   9m45s   3 OK / 3     True    
workflow                          http://workflow.custom-functions.10.107.55.54.sslip.io                          workflow-00001                          8s      3 OK / 3     True           

Submit a request with a plain JSON object

The service based on the JSON workflow definition can be accessed by sending a request to <URL>/plainJsonKnativeFunction (replace <URL> with the URL returned by the above command) with the following content:

{
  "name": "Kogito"
}

Complete curl command can be found below (replace <URL> with the URL for the workflow service returned by the kn service list command):

curl -X 'POST' \
  '<URL>/plainJsonKnativeFunction' \
  -H 'accept: */*' \
  -H 'Content-Type: application/json' \
  -d '{ "name": "Kogito" }'

Should return something like this (“id” will change):

{"id":"87cf8275-782d-4e0b-a9be-a95f95c9c190","workflowdata":{"name":"Kogito","greeting":"Greetings from Serverless Workflow, Kogito"}}

Submit a request with a CloudEvent

The service based on the JSON workflow definition can be accessed by sending a request to <URL>/cloudEventKnativeFunction (replace <URL> with the URL returned by the above command) with the following content:

{
  "cloudevent": {
    "specversion": "1.0",
    "source": "org.acme.source",
    "type": "test",
    "data": {
      "value": "test data"
    }
  }
}

[NOTE]

Kogito Serverless Workflow generates a CloudEvent ID based on source and the workflow instance ID. In case an ID is set, Kogito Serverless Workflow will ignore it and use a generated one.

Complete curl command can be found below (replace <URL> with the URL for the workflow service returned by the kn service list command):

curl -X 'POST' \
  '<URL>/cloudEventKnativeFunction' \
  -H 'accept: */*' \
  -H 'Content-Type: application/json' \
  -d '{ "cloudevent": { "specversion": "1.0", "source": "org.acme.source", "type": "test", "data": { "value": "test data"} }}'

Should return something like this (“id” will change):

{"id":"8dc00353-c1c6-45e9-845d-e9188d103f50","workflowdata":{"id":"response-of-org.acme.source_8dc00353-c1c6-45e9-845d-e9188d103f50","specversion":"1.0","source":"cloudEventFunction","type":"annotated"}}

Cleaning the example

To remove the installed services from your minikube installation you can use the following command:

# Note: this command might take some seconds.
kubectl delete namespace custom-functions