arazzo: 1.0.1
info:
  title: Amazon Neptune Bulk Load Job Lifecycle
  summary: Start a bulk load via the Loader endpoint, verify it appears in the job list, and poll its status.
  description: >-
    Drives the full lifecycle of a Neptune bulk load through the dedicated Loader
    REST endpoint. The workflow starts a bulk load job from Amazon S3, lists the
    recent load job ids to confirm the new job is tracked, and then polls the job
    status on a loop until the overall status reaches LOAD_COMPLETED. A poll loop
    handles LOAD_IN_PROGRESS. 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: neptuneLoaderApi
  url: ../openapi/amazon-neptune-loader-openapi.yml
  type: openapi
workflows:
- workflowId: loader-job-lifecycle
  summary: Start a load, confirm it is listed, and poll its status to completion.
  description: >-
    Starts a bulk load job, lists recent load ids to verify it is tracked, and
    polls the job status until it completes.
  inputs:
    type: object
    required:
    - source
    - format
    - iamRoleArn
    - region
    properties:
      source:
        type: string
        description: The S3 URI of the data files or folders to load.
      format:
        type: string
        description: The data format (csv, opencypher, ntriples, nquads, rdfxml, turtle).
      iamRoleArn:
        type: string
        description: The ARN of the IAM role with S3 access.
      region:
        type: string
        description: The AWS Region of the S3 bucket.
      failOnError:
        type: string
        description: Whether to stop the load on error (TRUE or FALSE).
  steps:
  - stepId: startLoad
    description: >-
      Start a bulk load job from the supplied S3 source and capture the load id.
    operationId: startBulkLoadJob
    requestBody:
      contentType: application/json
      payload:
        source: $inputs.source
        format: $inputs.format
        iamRoleArn: $inputs.iamRoleArn
        region: $inputs.region
        failOnError: $inputs.failOnError
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      loadId: $response.body#/payload/loadId
  - stepId: listJobs
    description: >-
      List the recent bulk load job ids, including queued loads, to confirm the
      new job is being tracked by Neptune.
    operationId: listBulkLoadJobs
    parameters:
    - name: includeQueuedLoads
      in: query
      value: true
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      loadIds: $response.body#/payload/loadIds
  - stepId: pollStatus
    description: >-
      Poll the load job status with detailed feed counts. Repeat while the
      overall status is LOAD_IN_PROGRESS and finish once it is LOAD_COMPLETED.
    operationId: getBulkLoadJobStatus
    parameters:
    - name: loadId
      in: path
      value: $steps.startLoad.outputs.loadId
    - name: details
      in: query
      value: true
    - name: errors
      in: query
      value: true
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      overallStatus: $response.body#/payload/overallStatus/status
      parsingErrors: $response.body#/payload/overallStatus/parsingErrors
    onSuccess:
    - name: loadStillRunning
      type: retry
      retryAfter: 30
      retryLimit: 60
      criteria:
      - context: $response.body
        condition: $.payload.overallStatus.status == "LOAD_IN_PROGRESS"
        type: jsonpath
    - name: loadComplete
      type: end
      criteria:
      - context: $response.body
        condition: $.payload.overallStatus.status == "LOAD_COMPLETED"
        type: jsonpath
  outputs:
    loadId: $steps.startLoad.outputs.loadId
    loadIds: $steps.listJobs.outputs.loadIds
    overallStatus: $steps.pollStatus.outputs.overallStatus