def execute()

in gcpdiag/runbook/dataproc/spark_job_failures.py [0:0]
57 lines of code
6 McCabe index (conditional complexity)

  def execute(self):
    """Verify job exists in Dataproc UI."""
    project = crm.get_project(op.get(flags.PROJECT_ID))

    if not op.get(flags.JOB_ID) or not op.get(flags.REGION):
      op.add_failed(
          project,
          reason=op.prep_msg(
              op.FAILURE_REASON,
              project_id=project,
              job_id=op.get(flags.JOB_ID),
              cluster_name=op.get(flags.CLUSTER_NAME),
          ),
          remediation=op.prep_msg(op.FAILURE_REMEDIATION),
      )
      return

    # uses the API to get the cluster information from the job id
    try:
      job = dataproc.get_job_by_jobid(project_id=op.get(flags.PROJECT_ID),
                                      region=op.get(flags.REGION),
                                      job_id=op.get(flags.JOB_ID))
    except (AttributeError, GcpApiError, IndexError, TypeError,
            ValueError) as e:
      op.put(flags.JOB_EXIST, 'false')
      op.add_failed(
          project,
          reason=op.prep_msg(
              op.FAILURE_REASON,
              project_id=project,
              job_id=op.get(flags.JOB_ID),
              cluster_name=op.get(flags.CLUSTER_NAME),
              error=e,
          ),
          remediation=op.prep_msg(op.FAILURE_REMEDIATION),
      )
      return

    # Start date is the date for when the job was running
    start_time = datetime.strptime(
        job.status_history['RUNNING'],
        '%Y-%m-%dT%H:%M:%S.%fZ').replace(tzinfo=timezone.utc)

    # End date is the start date + 7 days
    end_time = start_time + timedelta(days=7)

    # Saving cluster parameters
    op.put(flags.START_TIME, start_time)
    op.info(f'Start time utc:{start_time}')
    op.put(flags.END_TIME, end_time)
    op.info(f'End time utc:{end_time}')
    op.put(flags.CLUSTER_UUID, job.cluster_uuid)
    op.put(flags.CLUSTER_NAME, job.cluster_name)

    if check_datetime_gap(op.get(flags.START_TIME), op.get(flags.END_TIME), 30):
      op.put(flags.JOB_OLDER_THAN_30_DAYS, True)
    else:
      op.put(flags.JOB_OLDER_THAN_30_DAYS, False)

    op.add_ok(
        project,
        reason=op.prep_msg(
            op.SUCCESS_REASON,
            project_id=project,
            job_id=op.get(flags.JOB_ID),
            cluster_name=op.get(flags.CLUSTER_NAME),
        ),
    )
    return