arazzo: 1.0.1
info:
  title: Amazon EMR Launch a Spark Cluster
  summary: Create and start a new EMR cluster pre-configured to run Apache Spark.
  description: >-
    Launches a managed Amazon EMR cluster (job flow) with the Spark application
    installed so the cluster is ready to run large-scale distributed data
    processing and machine learning workloads. The workflow calls RunJobFlow
    with the supplied cluster name, instance configuration, and release label,
    requests the Spark application, and returns the new cluster's JobFlowId.
    Every step spells out its request inline, including the AWS JSON protocol
    X-Amz-Target header, so the flow can be read and executed without opening
    the underlying OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: emrApi
  url: ../openapi/amazon-emr-openapi.yml
  type: openapi
workflows:
- workflowId: run-spark-cluster
  summary: Run a new EMR cluster with the Spark application installed.
  description: >-
    Creates and starts a new EMR cluster using the provided instance
    configuration and release label, installing Apache Spark, and returns the
    identifier of the newly created cluster.
  inputs:
    type: object
    required:
    - name
    - instances
    - releaseLabel
    properties:
      name:
        type: string
        description: The name of the cluster to create.
      instances:
        type: object
        description: The instance configuration for the cluster (master/core/task layout).
      releaseLabel:
        type: string
        description: The Amazon EMR release label (e.g. emr-6.10.0).
  steps:
  - stepId: launchCluster
    description: >-
      Create and start a new EMR cluster with the Spark application using the
      supplied name, instance configuration, and release label.
    operationId: RunJobFlow
    parameters:
    - name: X-Amz-Target
      in: header
      value: ElasticMapReduce.RunJobFlow
    requestBody:
      contentType: application/json
      payload:
        Name: $inputs.name
        Instances: $inputs.instances
        ReleaseLabel: $inputs.releaseLabel
        Applications:
        - Name: Spark
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      jobFlowId: $response.body#/JobFlowId
  outputs:
    jobFlowId: $steps.launchCluster.outputs.jobFlowId