DataHub Decommission a Dataset

arazzo: 1.0.1
info:
  title: DataHub Decommission a Dataset
  summary: Confirm a dataset, check it has no downstream dependents, then soft delete it from the metadata graph.
  description: >-
    Safely retiring a dataset requires checking that nothing depends on it
    first. This workflow confirms the dataset exists, queries the relationship
    graph for outgoing DownstreamOf edges to detect any downstream dependents,
    and branches: when downstream dependents are found it stops, and when none are
    found it performs a soft delete that marks the entity as removed while
    preserving its metadata. 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: entitiesApi
  url: ../openapi/datahub-entities-api-openapi.yml
  type: openapi
- name: relationshipsApi
  url: ../openapi/datahub-relationships-api-openapi.yml
  type: openapi
workflows:
- workflowId: decommission-dataset
  summary: Soft delete a dataset only when it has no downstream dependents.
  description: >-
    Confirms a dataset exists, checks its DownstreamOf relationships, and soft
    deletes the entity only when no downstream dependents are present.
  inputs:
    type: object
    required:
    - token
    - entityUrn
    properties:
      token:
        type: string
        description: DataHub personal access token passed as a Bearer token.
      entityUrn:
        type: string
        description: The dataset URN to decommission.
  steps:
  - stepId: confirmDataset
    description: >-
      Retrieve the latest aspects for the dataset URN to confirm the entity
      exists before attempting to decommission it.
    operationId: getEntityLatestAspects
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    - name: urns
      in: query
      value: $inputs.entityUrn
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      confirmedUrn: $response.body#/0/entityUrn
  - stepId: checkDependents
    description: >-
      Query the relationship graph for outgoing DownstreamOf edges to detect
      any datasets that depend on this one before deleting it.
    operationId: getRelationships
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    - name: urn
      in: query
      value: $steps.confirmDataset.outputs.confirmedUrn
    - name: relationshipTypes
      in: query
      value:
      - DownstreamOf
    - name: direction
      in: query
      value: OUTGOING
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      total: $response.body#/total
    onSuccess:
    - name: noDependents
      type: goto
      stepId: softDelete
      criteria:
      - context: $response.body
        condition: $.relationships.length == 0
        type: jsonpath
    - name: hasDependents
      type: end
      criteria:
      - context: $response.body
        condition: $.relationships.length > 0
        type: jsonpath
  - stepId: softDelete
    description: >-
      Soft delete the dataset, marking the entity as removed while preserving
      its metadata, since no downstream dependents were found.
    operationId: deleteEntities
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    - name: urns
      in: query
      value:
      - $inputs.entityUrn
    - name: soft
      in: query
      value: true
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      deleteStatus: $statusCode
  outputs:
    dependentCount: $steps.checkDependents.outputs.total
    deleteStatus: $steps.softDelete.outputs.deleteStatus