arazzo: 1.0.1
info:
  title: Databricks Create Job and Trigger First Run
  summary: Create a new job from settings, then immediately trigger its first run.
  description: >-
    Defines a new Databricks job from a supplied settings payload and then
    triggers an immediate run of that job, returning the new run_id so callers
    can track execution. The job_id produced by the create call is reused for
    the run-now request. 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: databricksApi
  url: ../openapi/databricks-openapi.yml
  type: openapi
workflows:
- workflowId: create-and-run-job
  summary: Create a job and trigger its first run.
  description: >-
    Creates a job with the given name and task list, then immediately submits a
    run-now request for the newly created job.
  inputs:
    type: object
    required:
    - name
    - tasks
    properties:
      name:
        type: string
        description: The display name for the new job.
      tasks:
        type: array
        description: The list of task definitions that make up the job.
      job_clusters:
        type: array
        description: Optional list of shared job cluster definitions.
      notebook_params:
        type: object
        description: Optional notebook parameters to pass on the first run.
  steps:
  - stepId: createJob
    description: >-
      Create the job from the supplied name and tasks. Returns the canonical
      job_id used to trigger the run.
    operationId: createJob
    requestBody:
      contentType: application/json
      payload:
        name: $inputs.name
        tasks: $inputs.tasks
        job_clusters: $inputs.job_clusters
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      jobId: $response.body#/job_id
  - stepId: triggerFirstRun
    description: >-
      Trigger the first run of the newly created job, optionally passing
      notebook parameters. Returns the run_id of the triggered run.
    operationId: runJobNow
    requestBody:
      contentType: application/json
      payload:
        job_id: $steps.createJob.outputs.jobId
        notebook_params: $inputs.notebook_params
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      runId: $response.body#/run_id
      numberInJob: $response.body#/number_in_job
  outputs:
    jobId: $steps.createJob.outputs.jobId
    runId: $steps.triggerFirstRun.outputs.runId