Bright Data Rerun a Failed Snapshot and Monitor It

arazzo: 1.0.1
info:
  title: Bright Data Rerun a Failed Snapshot and Monitor It
  summary: Inspect an existing snapshot, rerun it when it failed, and poll the new snapshot.
  description: >-
    A recovery pattern for the Web Scraper API. The workflow reads the progress
    of an existing snapshot and, when it did not complete successfully, reruns
    it to produce a fresh snapshot id, then polls the new snapshot until it
    reaches a terminal status. 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: webScraperApi
  url: ../openapi/bright-data-web-scraper-api-openapi.yml
  type: openapi
workflows:
- workflowId: rerun-and-monitor-snapshot
  summary: Check a snapshot, rerun it when not ready, and poll the new snapshot.
  description: >-
    Reads the current progress of a snapshot, and when its status is not ready
    triggers a rerun, then polls the freshly created snapshot until it finishes.
  inputs:
    type: object
    required:
    - apiToken
    - snapshotId
    properties:
      apiToken:
        type: string
        description: Bright Data API token used as a Bearer credential.
      snapshotId:
        type: string
        description: Identifier of the existing snapshot to inspect and rerun.
  steps:
  - stepId: checkProgress
    description: >-
      Read the progress of the existing snapshot. A failed or cancelled status
      means the snapshot should be rerun.
    operationId: getScrapeProgress
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiToken"
    - name: snapshot_id
      in: path
      value: $inputs.snapshotId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      status: $response.body#/status
    onSuccess:
    - name: needsRerun
      type: goto
      stepId: rerunSnapshot
      criteria:
      - context: $response.body
        condition: $.status != "ready"
        type: jsonpath
    - name: alreadyReady
      type: end
      criteria:
      - context: $response.body
        condition: $.status == "ready"
        type: jsonpath
  - stepId: rerunSnapshot
    description: >-
      Rerun the snapshot, producing a new snapshot id that reprocesses the
      original inputs.
    operationId: rerunSnapshot
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiToken"
    - name: snapshot_id
      in: path
      value: $inputs.snapshotId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      newSnapshotId: $response.body#/snapshot_id
  - stepId: pollRerun
    description: >-
      Poll the rerun snapshot until it reaches a terminal status. A ready status
      means the rerun produced rows.
    operationId: getScrapeProgress
    parameters:
    - name: Authorization
      in: header
      value: "Bearer $inputs.apiToken"
    - name: snapshot_id
      in: path
      value: $steps.rerunSnapshot.outputs.newSnapshotId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      status: $response.body#/status
      records: $response.body#/records
    onSuccess:
    - name: keepPolling
      type: goto
      stepId: pollRerun
      criteria:
      - context: $response.body
        condition: $.status != "ready" && $.status != "failed" && $.status != "cancelled"
        type: jsonpath
  outputs:
    originalStatus: $steps.checkProgress.outputs.status
    newSnapshotId: $steps.rerunSnapshot.outputs.newSnapshotId
    rerunStatus: $steps.pollRerun.outputs.status