arazzo: 1.0.1
info:
  title: Red Hat OpenShift Provision Cluster and Add Machine Pool
  summary: Pick a version and cloud provider, create a cluster, poll until ready, then add a machine pool.
  description: >-
    A full provisioning flow for the Red Hat OpenShift Cluster Manager
    (clusters_mgmt v1). The workflow resolves an available OpenShift version and
    a target cloud provider, creates a cluster, polls the cluster until it
    leaves its installing state and becomes ready, then attaches an additional
    machine pool of compute nodes. Each step inlines its bearer token, request
    body, documented success criteria, and outputs so the flow is fully
    self-describing.
  version: 1.0.0
sourceDescriptions:
- name: openshiftClusterManagerApi
  url: ../openapi/red-hat-openshift-cluster-manager-openapi.yml
  type: openapi
workflows:
- workflowId: provision-cluster
  summary: Resolve version and provider, create a cluster, wait for ready, add a machine pool.
  description: >-
    Lists available OpenShift versions and cloud providers, creates a cluster
    using the supplied name and region, polls the cluster until its state is
    ready, and then creates an additional machine pool.
  inputs:
    type: object
    required:
    - token
    - name
    - regionId
    properties:
      token:
        type: string
        description: OCM bearer token for the Cluster Manager API.
      name:
        type: string
        description: The name for the new cluster.
      display_name:
        type: string
        description: An optional human-friendly display name for the cluster.
      regionId:
        type: string
        description: The cloud provider region id to deploy into.
      multi_az:
        type: boolean
        description: Whether to spread the cluster across multiple availability zones.
      compute:
        type: integer
        description: The number of compute (worker) nodes for the cluster.
      instance_type:
        type: string
        description: The instance type for the additional machine pool.
      replicas:
        type: integer
        description: The replica count for the additional machine pool.
  steps:
  - stepId: resolveVersion
    description: >-
      List available OpenShift versions and select the first enabled entry to
      install on the new cluster.
    operationId: listVersions
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      versionId: $response.body#/items/0/id
  - stepId: resolveProvider
    description: >-
      List supported cloud providers and select the first available provider id
      to host the cluster.
    operationId: listCloudProviders
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      cloudProviderId: $response.body#/items/0/id
  - stepId: createCluster
    description: >-
      Create the cluster using the resolved version, provider, and supplied
      region. The API returns a cluster record whose state begins as installing.
    operationId: createCluster
    parameters:
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    requestBody:
      contentType: application/json
      payload:
        name: $inputs.name
        display_name: $inputs.display_name
        cloud_provider:
          id: $steps.resolveProvider.outputs.cloudProviderId
        region:
          id: $inputs.regionId
        multi_az: $inputs.multi_az
        nodes:
          compute: $inputs.compute
        version:
          id: $steps.resolveVersion.outputs.versionId
    successCriteria:
    - condition: $statusCode == 201
    outputs:
      clusterId: $response.body#/id
      state: $response.body#/state
  - stepId: pollCluster
    description: >-
      Retrieve the cluster and re-poll while it is still installing; proceed to
      add a machine pool once the cluster state becomes ready.
    operationId: getCluster
    parameters:
    - name: cluster_id
      in: path
      value: $steps.createCluster.outputs.clusterId
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      state: $response.body#/state
      apiUrl: $response.body#/api/url
    onSuccess:
    - name: stillInstalling
      type: goto
      stepId: pollCluster
      criteria:
      - context: $response.body
        condition: $.state == 'installing' || $.state == 'pending' || $.state == 'waiting'
        type: jsonpath
    - name: ready
      type: goto
      stepId: addMachinePool
      criteria:
      - context: $response.body
        condition: $.state == 'ready'
        type: jsonpath
  - stepId: addMachinePool
    description: >-
      Create an additional machine pool of compute nodes on the now-ready
      cluster.
    operationId: createMachinePool
    parameters:
    - name: cluster_id
      in: path
      value: $steps.createCluster.outputs.clusterId
    - name: Authorization
      in: header
      value: Bearer $inputs.token
    requestBody:
      contentType: application/json
      payload:
        instance_type: $inputs.instance_type
        replicas: $inputs.replicas
    successCriteria:
    - condition: $statusCode == 201
    outputs:
      machinePoolId: $response.body#/id
  outputs:
    clusterId: $steps.createCluster.outputs.clusterId
    clusterState: $steps.pollCluster.outputs.state
    machinePoolId: $steps.addMachinePool.outputs.machinePoolId