如何以编程方式检查Spark step状态(在EMR集群上提交)？

import boto3

# 创建EMR客户端
emr_client = boto3.client('emr', region_name='your_region')

# 定义EMR集群和步骤的标识符
cluster_id = 'your_cluster_id'
step_id = 'your_step_id'

# 调用DescribeStep API获取步骤的详细信息
response = emr_client.describe_step(
    ClusterId=cluster_id,
    StepId=step_id
)

# 解析API响应获取步骤的状态
step_status = response['Step']['Status']['State']

# 根据步骤的状态采取相应的操作
if step_status == 'COMPLETED':
    # 获取步骤的输出日志
    step_logs = response['Step']['Status']['Timeline']['CreationDateTime']
    print('Step completed successfully. Logs:', step_logs)
elif step_status == 'FAILED':
    print('Step execution failed.')
else:
    print('Step is still running or pending.')