arazzo: 1.0.1
info:
  title: Amazon Kendra Provision Index and Start First Sync
  summary: Create an index, wait until it is active, attach a data source, and kick off the first sync job.
  description: >-
    Stands up a brand new Amazon Kendra search index from scratch and gets it
    ingesting content. The workflow creates the index, polls DescribeIndex until
    the index leaves the CREATING state and reaches ACTIVE, then registers a data
    source connector against the active index and immediately starts a
    synchronization job so the connector begins crawling the repository. Each
    step inlines its request, including the AWS JSON protocol X-Amz-Target
    header, so the flow can be read and executed without opening the underlying
    OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: kendraApi
  url: ../openapi/amazon-kendra-openapi.yml
  type: openapi
workflows:
- workflowId: provision-index-and-sync
  summary: Create an index, wait for ACTIVE, add a data source, and start a sync job.
  description: >-
    Creates a new Kendra index, polls until it is ACTIVE, creates a data source
    connector on the active index, and starts the first synchronization job.
  inputs:
    type: object
    required:
    - indexName
    - indexRoleArn
    - dataSourceName
    - dataSourceType
    properties:
      indexName:
        type: string
        description: The name for the new search index.
      indexRoleArn:
        type: string
        description: The IAM role ARN that gives Kendra permission to manage the index.
      indexEdition:
        type: string
        description: The index edition, DEVELOPER_EDITION or ENTERPRISE_EDITION.
      dataSourceName:
        type: string
        description: The name for the data source connector.
      dataSourceType:
        type: string
        description: The data source connector type, such as S3 or SHAREPOINT.
      dataSourceRoleArn:
        type: string
        description: The IAM role ARN for the data source connector.
      dataSourceSchedule:
        type: string
        description: An optional cron schedule for recurring synchronization.
  steps:
  - stepId: createIndex
    description: >-
      Create the new Amazon Kendra index. The index begins life in the CREATING
      state and is not queryable until it becomes ACTIVE.
    operationId: CreateIndex
    parameters:
    - name: X-Amz-Target
      in: header
      value: AWSKendraFrontendService.CreateIndex
    requestBody:
      contentType: application/json
      payload:
        Name: $inputs.indexName
        RoleArn: $inputs.indexRoleArn
        Edition: $inputs.indexEdition
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      indexId: $response.body#/Id
  - stepId: waitForIndexActive
    description: >-
      Poll the index until provisioning completes. The index reports a Status of
      CREATING while it is being built and ACTIVE once it can serve queries and
      accept documents.
    operationId: DescribeIndex
    parameters:
    - name: IndexId
      in: path
      value: $steps.createIndex.outputs.indexId
    - name: X-Amz-Target
      in: header
      value: AWSKendraFrontendService.DescribeIndex
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      indexStatus: $response.body#/Status
    onSuccess:
    - name: indexReady
      type: goto
      stepId: createDataSource
      criteria:
      - context: $response.body
        condition: $.Status == "ACTIVE"
        type: jsonpath
    onFailure:
    - name: retryDescribe
      type: retry
      stepId: waitForIndexActive
      retryAfter: 30
      retryLimit: 20
      criteria:
      - context: $response.body
        condition: $.Status == "CREATING"
        type: jsonpath
  - stepId: createDataSource
    description: >-
      Register a data source connector against the now-active index so Kendra
      knows where to crawl content from.
    operationId: CreateDataSource
    parameters:
    - name: IndexId
      in: path
      value: $steps.createIndex.outputs.indexId
    - name: X-Amz-Target
      in: header
      value: AWSKendraFrontendService.CreateDataSource
    requestBody:
      contentType: application/json
      payload:
        Name: $inputs.dataSourceName
        Type: $inputs.dataSourceType
        RoleArn: $inputs.dataSourceRoleArn
        Schedule: $inputs.dataSourceSchedule
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      dataSourceId: $response.body#/Id
  - stepId: startSyncJob
    description: >-
      Start the first synchronization job so the connector begins crawling and
      indexing documents from the configured repository.
    operationId: StartDataSourceSyncJob
    parameters:
    - name: IndexId
      in: path
      value: $steps.createIndex.outputs.indexId
    - name: DataSourceId
      in: path
      value: $steps.createDataSource.outputs.dataSourceId
    - name: X-Amz-Target
      in: header
      value: AWSKendraFrontendService.StartDataSourceSyncJob
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      executionId: $response.body#/ExecutionId
  outputs:
    indexId: $steps.createIndex.outputs.indexId
    dataSourceId: $steps.createDataSource.outputs.dataSourceId
    executionId: $steps.startSyncJob.outputs.executionId