使用 Python SDK

使用 iaesdk 的代码样本

安装 Python SDK 后，开始使用该 SDK，包括将凭证寻源到 IBM Analytics Engine 服务，调用该服务，然后发出不同的集群命令，如以下样本代码片段中所示。 代码示例是针对 Python 3.7编写的。

除了本主题中的样本代码片段外，您还可以使用 IBM Analytics Engine V3 API 参考 中的 Python 代码样本。

以下代码样本显示了如何执行以下操作:

• 向 IBM Analytics Engine 服务认证并构建服务客户机:

  from iaesdk import IbmAnalyticsEngineApiV3
  from ibm_cloud_sdk_core.authenticators import IAMAuthenticator
  import json
  
  # Constants for IBM Analytics Engine values
  IAM_API_KEY = "{apikey}" # eg "W00YiRnLW4a3fTjMB-odB-2ySfTrFBIQQWanc--P3byk"
  IAE_ENDPOINT_URL = "{url}" # Current list avaiable at https://cloud.ibm.com/apidocs/ibm-analytics-engine#service-endpoints
  const API_AUTH_URL = "{api auth url}"  # "https://iam.cloud.ibm.com/identity/token"
  
  # Create an IAM authenticator.
  authenticator = IAMAuthenticator(apikey=IAM_API_KEY, url=API_AUTH_URL)
  
  # Construct the service client.
  ibm_analytics_engine_api_service = IbmAnalyticsEngineApiV3(authenticator=authenticator)
  
  # Set our custom service URL
  ibm_analytics_engine_api_service.set_service_url(IAE_ENDPOINT_URL)
  
  # Service operations can now be invoked using the "ibm_analytics_engine_api_service" variable.
  
  

• 检索单个实例的详细信息:

  get_instance(self,
      instance_id: str,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  instance = ibm_analytics_engine_api_service.get_instance(
  instance_id='dc0e9889-eab2-4t9e-9441-566209499546'
  ).get_result()
  print(json.dumps(instance, indent=2))
  

• 在给定的无服务器 Spark 实例上部署 Spark 应用程序:

  create_application(self,
      instance_id: str,
      *,
      application_details: 'ApplicationRequestApplicationDetails' = None,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  application_request_application_details_model = {
    'application': '/opt/ibm/spark/examples/src/main/python/wordcount.py',
    'arguments': ['/opt/ibm/spark/examples/src/main/resources/people.txt']
    }
  create_application_response = ibm_analytics_engine_api_service.create_application(
    instance_id='dc0e9889-eab2-4t9e-9441-566209499546',
    application_details=application_request_application_details_model
  ).get_result()
  print(json.dumps(create_application_response, indent=2))
  

• 检索在给定实例上运行的所有 Spark 应用程序:

  list_applications(self,
      instance_id: str,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  list_applications_response = ibm_analytics_engine_api_service.list_applications(
    instance_id='dc0e9889-eab2-4t9e-9441-566209499546'
    ).get_result()
  print(json.dumps(list_applications_response, indent=2))
  

• 检索给定 Spark 应用程序的详细信息:

  get_application(self,
      instance_id: str,
      application_id: str,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  get_application_response = ibm_analytics_engine_api_service.get_application(
    instance_id='dc0e9889-eab2-4t9e-9441-566209499546',
    application_id='db933645-0b68-4dcb-80d8-7b71a6c8e542'
    ).get_result()
  print(json.dumps(get_application_response, indent=2))
  

• 停止由 app_id 标识标识的正在运行的应用程序。 这是一个幂等的操作。 如果请求的应用程序已停止或完成，那么不执行任何操作。

  delete_application(self,
      instance_id: str,
      application_id: str,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  delete_application_by_id_response = ibm_analytics_engine_api_service.delete_application(
    instance_id='dc0e9889-eab2-4t9e-9441-566209499546',
    application_id='db933645-0b68-4dcb-80d8-7b71a6c8e542'
    ).get_result()
  print(json.dumps(delete_application_by_id_response, indent=2))
  

• 返回由 app_id 标识标识的应用程序的状态:

  get_application_state(self,
      instance_id: str,
      application_id: str,
      **kwargs
  ) -> DetailedResponse
  

  示例请求：

  get_application_state_response = ibm_analytics_engine_api_service.get_application_state(
    instance_id='dc0e9889-eab2-4t9e-9441-566209499546',
    application_id='db933645-0b68-4dcb-80d8-7b71a6c8e542'
    ).get_result()
  print(json.dumps(get_application_state_response, indent=2))