arazzo: 1.0.1
info:
  title: Amazon Data Pipeline Provision and Activate
  summary: Create an empty pipeline, populate its definition, activate it, and confirm its state.
  description: >-
    The canonical end-to-end provisioning flow for AWS Data Pipeline. An empty
    pipeline is created with a caller-supplied unique id, the pipeline
    definition (objects such as schedules, defaults, and activities) is written
    with PutPipelineDefinition, the pipeline is activated to begin processing
    tasks, and DescribePipelines confirms the resulting pipeline state. Every
    step spells out its request inline so the flow can be read and executed
    without opening the underlying OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: dataPipelineApi
  url: ../openapi/amazon-data-pipeline-openapi.yml
  type: openapi
workflows:
- workflowId: provision-and-activate-pipeline
  summary: Create a pipeline, set its definition, activate it, and confirm state.
  description: >-
    Chains createPipeline into putPipelineDefinition, then activatePipeline, and
    finally describePipelines so the newly provisioned pipeline is fully running
    and its state is observed in a single pass.
  inputs:
    type: object
    required:
    - name
    - uniqueId
    - pipelineObjects
    properties:
      name:
        type: string
        description: The human-readable name of the pipeline.
      uniqueId:
        type: string
        description: An idempotency token that prevents duplicate pipeline creation.
      description:
        type: string
        description: An optional description of the pipeline.
      pipelineObjects:
        type: array
        description: The pipeline objects (schedules, defaults, activities) that make up the definition.
        items:
          type: object
      startTimestamp:
        type: string
        description: The optional date-time at which the pipeline should begin processing.
  steps:
  - stepId: createPipeline
    description: >-
      Create a new, empty pipeline shell using the supplied name and idempotent
      unique id.
    operationId: createPipeline
    requestBody:
      contentType: application/json
      payload:
        name: $inputs.name
        uniqueId: $inputs.uniqueId
        description: $inputs.description
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      pipelineId: $response.body#/pipelineId
  - stepId: putDefinition
    description: >-
      Populate the freshly created pipeline with its objects (schedules,
      defaults, activities, preconditions).
    operationId: putPipelineDefinition
    requestBody:
      contentType: application/json
      payload:
        pipelineId: $steps.createPipeline.outputs.pipelineId
        pipelineObjects: $inputs.pipelineObjects
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      errored: $response.body#/errored
      validationErrors: $response.body#/validationErrors
  - stepId: activatePipeline
    description: >-
      Validate and start the pipeline so it begins processing its tasks on the
      configured schedule.
    operationId: activatePipeline
    requestBody:
      contentType: application/json
      payload:
        pipelineId: $steps.createPipeline.outputs.pipelineId
        startTimestamp: $inputs.startTimestamp
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      activated: $statusCode
  - stepId: confirmState
    description: >-
      Read back the pipeline metadata to confirm the pipeline exists and observe
      its current state after activation.
    operationId: describePipelines
    requestBody:
      contentType: application/json
      payload:
        pipelineIds:
        - $steps.createPipeline.outputs.pipelineId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      pipelineState: $response.body#/pipelineDescriptionList/0/pipelineState
  outputs:
    pipelineId: $steps.createPipeline.outputs.pipelineId
    pipelineState: $steps.confirmState.outputs.pipelineState