arazzo: 1.0.1
info:
  title: Amazon RDS Delete an Aurora DB Cluster
  summary: Snapshot member instances, delete them, then delete the Aurora cluster.
  description: >-
    Tears down an Aurora DB cluster member and the cluster itself with a final
    cluster snapshot retained. The workflow confirms the cluster is available,
    deletes the supplied member instance while skipping its individual snapshot,
    waits for the instance to disappear, and then deletes the cluster while
    retaining a final cluster snapshot. 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: amazonRdsApi
  url: ../openapi/amazon-rds-openapi.yml
  type: openapi
workflows:
- workflowId: delete-aurora-cluster
  summary: Delete an Aurora cluster member then the cluster with a final snapshot.
  description: >-
    Confirms the cluster is available, deletes the member instance, polls until
    the instance is gone, then deletes the cluster retaining a final snapshot.
  inputs:
    type: object
    required:
    - dbClusterIdentifier
    - dbInstanceIdentifier
    - finalDbSnapshotIdentifier
    properties:
      dbClusterIdentifier:
        type: string
        description: The identifier of the Aurora DB cluster to delete.
      dbInstanceIdentifier:
        type: string
        description: The identifier of the cluster member instance to delete first.
      finalDbSnapshotIdentifier:
        type: string
        description: The identifier for the final cluster snapshot taken during deletion.
  steps:
  - stepId: confirmCluster
    description: >-
      Describe the Aurora DB cluster to confirm it is available before tearing
      it down.
    operationId: describeDBClusters
    parameters:
    - name: DBClusterIdentifier
      in: query
      value: $inputs.dbClusterIdentifier
    successCriteria:
    - condition: $statusCode == 200
    - context: $response.body
      condition: $.dBClusters[0].status == 'available'
      type: jsonpath
    outputs:
      clusterStatus: $response.body#/dBClusters/0/status
  - stepId: deleteMember
    description: >-
      Delete the cluster member instance, skipping a per-instance final
      snapshot since a cluster-level snapshot is taken later.
    operationId: deleteDBInstance
    parameters:
    - name: DBInstanceIdentifier
      in: query
      value: $inputs.dbInstanceIdentifier
    - name: SkipFinalSnapshot
      in: query
      value: true
    successCriteria:
    - condition: $statusCode == 200
  - stepId: pollMemberGone
    description: >-
      Describe the member instance to track its teardown. While it is still
      deleting the workflow loops back to poll again; once it reports deleting
      it proceeds to remove the cluster.
    operationId: describeDBInstances
    parameters:
    - name: DBInstanceIdentifier
      in: query
      value: $inputs.dbInstanceIdentifier
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      memberStatus: $response.body#/dBInstances/0/dBInstanceStatus
    onSuccess:
    - name: stillDeleting
      type: goto
      stepId: deleteCluster
      criteria:
      - context: $response.body
        condition: $.dBInstances[0].dBInstanceStatus == 'deleting'
        type: jsonpath
  - stepId: deleteCluster
    description: >-
      Delete the Aurora DB cluster, retaining a final cluster snapshot rather
      than skipping it.
    operationId: deleteDBCluster
    parameters:
    - name: DBClusterIdentifier
      in: query
      value: $inputs.dbClusterIdentifier
    - name: SkipFinalSnapshot
      in: query
      value: false
    - name: FinalDBSnapshotIdentifier
      in: query
      value: $inputs.finalDbSnapshotIdentifier
    successCriteria:
    - condition: $statusCode == 200
  outputs:
    clusterStatus: $steps.confirmCluster.outputs.clusterStatus
    memberStatus: $steps.pollMemberGone.outputs.memberStatus