class DataflowCreatePythonJobOperator(BaseOperator): """ Launching Cloud Dataflow jobs written in python. Note that both dataflow_default_options and options will be merged to specify pipeline execution parameter, and dataflow_default_options is expected to save high-level options, for instances, project and zone information, which apply to all dataflow operators in the DAG. This class is deprecated. Please use `providers.apache.beam.operators.beam.BeamRunPythonPipelineOperator`. .. seealso:: For more detail on job submission have a look at the reference: https://cloud.google.com/dataflow/pipelines/specifying-exec-params .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:DataflowCreatePythonJobOperator` :param py_file: Reference to the python dataflow pipeline file.py, e.g., /some/local/file/path/to/your/python/pipeline/file. (templated) :param job_name: The 'job_name' to use when executing the Dataflow job (templated). This ends up being set in the pipeline options, so any entry with key ``'jobName'`` or ``'job_name'`` in ``options`` will be overwritten. :param py_options: Additional python options, e.g., ["-m", "-v"]. :param dataflow_default_options: Map of default job options. :param options: Map of job specific options.The key must be a dictionary. The value can contain different types: * If the value is None, the single option - ``--key`` (without value) will be added. * If the value is False, this option will be skipped * If the value is True, the single option - ``--key`` (without value) will be added. * If the value is list, the many options will be added for each key. If the value is ``['A', 'B']`` and the key is ``key`` then the ``--key=A --key=B`` options will be left * Other value types will be replaced with the Python textual representation. When defining labels (``labels`` option), you can also provide a dictionary. :param py_interpreter: Python version of the beam pipeline. If None, this defaults to the python3. To track python versions supported by beam and related issues check: https://issues.apache.org/jira/browse/BEAM-1251 :param py_requirements: Additional python package(s) to install. If a value is passed to this parameter, a new virtual environment has been created with additional packages installed. You could also install the apache_beam package if it is not installed on your system or you want to use a different version. :param py_system_site_packages: Whether to include system_site_packages in your virtualenv. See virtualenv documentation for more information. This option is only relevant if the ``py_requirements`` parameter is not None. :param gcp_conn_id: The connection ID to use connecting to Google Cloud. :param project_id: Optional, the Google Cloud project ID in which to start a job. If set to None or missing, the default project_id from the Google Cloud connection is used. :param location: Job location. :param delegate_to: The account to impersonate using domain-wide delegation of authority, if any. For this to work, the service account making the request must have domain-wide delegation enabled. :param poll_sleep: The time in seconds to sleep between polling Google Cloud Platform for the dataflow job status while the job is in the JOB_STATE_RUNNING state. :param drain_pipeline: Optional, set to True if want to stop streaming job by draining it instead of canceling during killing task instance. See: https://cloud.google.com/dataflow/docs/guides/stopping-a-pipeline :param cancel_timeout: How long (in seconds) operator should wait for the pipeline to be successfully cancelled when task is being killed. :param wait_until_finished: (Optional) If True, wait for the end of pipeline execution before exiting. If False, only submits job. If None, default behavior. The default behavior depends on the type of pipeline: * for the streaming pipeline, wait for jobs to start, * for the batch pipeline, wait for the jobs to complete. .. warning:: You cannot call ``PipelineResult.wait_until_finish`` method in your pipeline code for the operator to work properly. i. e. you must use asynchronous execution. Otherwise, your pipeline will always wait until finished. For more information, look at: `Asynchronous execution <https://cloud.google.com/dataflow/docs/guides/specifying-exec-params#python_10>`__ The process of starting the Dataflow job in Airflow consists of two steps: * running a subprocess and reading the stderr/stderr log for the job id. * loop waiting for the end of the job ID from the previous step. This loop checks the status of the job. Step two is started just after step one has finished, so if you have wait_until_finished in your pipeline code, step two will not start until the process stops. When this process stops, steps two will run, but it will only execute one iteration as the job will be in a terminal state. If you in your pipeline do not call the wait_for_pipeline method but pass wait_until_finish=True to the operator, the second loop will wait for the job's terminal state. If you in your pipeline do not call the wait_for_pipeline method, and pass wait_until_finish=False to the operator, the second loop will check once is job not in terminal state and exit the loop. """ template_fields: Sequence[str] = ("options", "dataflow_default_options", "job_name", "py_file") def __init__( self, *, py_file: str, job_name: str = "{{task.task_id}}", dataflow_default_options: Optional[dict] = None, options: Optional[dict] = None, py_interpreter: str = "python3", py_options: Optional[List[str]] = None, py_requirements: Optional[List[str]] = None, py_system_site_packages: bool = False, project_id: Optional[str] = None, location: str = DEFAULT_DATAFLOW_LOCATION, gcp_conn_id: str = "google_cloud_default", delegate_to: Optional[str] = None, poll_sleep: int = 10, drain_pipeline: bool = False, cancel_timeout: Optional[int] = 10 * 60, wait_until_finished: Optional[bool] = None, **kwargs, ) -> None: # TODO: Remove one day warnings.warn( f"The `{self.__class__.__name__}` operator is deprecated, " "please use `providers.apache.beam.operators.beam.BeamRunPythonPipelineOperator` instead.", DeprecationWarning, stacklevel=2, ) super().__init__(**kwargs) self.py_file = py_file self.job_name = job_name self.py_options = py_options or [] self.dataflow_default_options = dataflow_default_options or {} self.options = options or {} self.options.setdefault("labels", {}).update({ "airflow-version": "v" + version.replace(".", "-").replace("+", "-") }) self.py_interpreter = py_interpreter self.py_requirements = py_requirements self.py_system_site_packages = py_system_site_packages self.project_id = project_id self.location = location self.gcp_conn_id = gcp_conn_id self.delegate_to = delegate_to self.poll_sleep = poll_sleep self.drain_pipeline = drain_pipeline self.cancel_timeout = cancel_timeout self.wait_until_finished = wait_until_finished self.job_id = None self.beam_hook: Optional[BeamHook] = None self.dataflow_hook: Optional[DataflowHook] = None def execute(self, context: 'Context'): """Execute the python dataflow job.""" self.beam_hook = BeamHook(runner=BeamRunnerType.DataflowRunner) self.dataflow_hook = DataflowHook( gcp_conn_id=self.gcp_conn_id, delegate_to=self.delegate_to, poll_sleep=self.poll_sleep, impersonation_chain=None, drain_pipeline=self.drain_pipeline, cancel_timeout=self.cancel_timeout, wait_until_finished=self.wait_until_finished, ) job_name = self.dataflow_hook.build_dataflow_job_name( job_name=self.job_name) pipeline_options = self.dataflow_default_options.copy() pipeline_options["job_name"] = job_name pipeline_options[ "project"] = self.project_id or self.dataflow_hook.project_id pipeline_options["region"] = self.location pipeline_options.update(self.options) # Convert argument names from lowerCamelCase to snake case. camel_to_snake = lambda name: re.sub( r"[A-Z]", lambda x: "_" + x.group(0).lower(), name) formatted_pipeline_options = { camel_to_snake(key): pipeline_options[key] for key in pipeline_options } def set_current_job_id(job_id): self.job_id = job_id process_line_callback = process_line_and_extract_dataflow_job_id_callback( on_new_job_id_callback=set_current_job_id) with ExitStack() as exit_stack: if self.py_file.lower().startswith("gs://"): gcs_hook = GCSHook(self.gcp_conn_id, self.delegate_to) tmp_gcs_file = exit_stack.enter_context( gcs_hook.provide_file(object_url=self.py_file)) self.py_file = tmp_gcs_file.name with self.dataflow_hook.provide_authorized_gcloud(): self.beam_hook.start_python_pipeline( variables=formatted_pipeline_options, py_file=self.py_file, py_options=self.py_options, py_interpreter=self.py_interpreter, py_requirements=self.py_requirements, py_system_site_packages=self.py_system_site_packages, process_line_callback=process_line_callback, ) self.dataflow_hook.wait_for_done( job_name=job_name, location=self.location, job_id=self.job_id, multiple_jobs=False, ) return {"job_id": self.job_id} def on_kill(self) -> None: self.log.info("On kill.") if self.job_id: self.dataflow_hook.cancel_job(job_id=self.job_id, project_id=self.project_id or self.dataflow_hook.project_id)
class DataflowCreateJavaJobOperator(BaseOperator): """ Start a Java Cloud Dataflow batch job. The parameters of the operation will be passed to the job. This class is deprecated. Please use `providers.apache.beam.operators.beam.BeamRunJavaPipelineOperator`. **Example**: :: default_args = { "owner": "airflow", "depends_on_past": False, "start_date": (2016, 8, 1), "email": ["*****@*****.**"], "email_on_failure": False, "email_on_retry": False, "retries": 1, "retry_delay": timedelta(minutes=30), "dataflow_default_options": { "project": "my-gcp-project", "zone": "us-central1-f", "stagingLocation": "gs://bucket/tmp/dataflow/staging/", }, } dag = DAG("test-dag", default_args=default_args) task = DataflowCreateJavaJobOperator( gcp_conn_id="gcp_default", task_id="normalize-cal", jar="{{var.value.gcp_dataflow_base}}pipeline-ingress-cal-normalize-1.0.jar", options={ "autoscalingAlgorithm": "BASIC", "maxNumWorkers": "50", "start": "{{ds}}", "partitionType": "DAY", }, dag=dag, ) .. seealso:: For more detail on job submission have a look at the reference: https://cloud.google.com/dataflow/pipelines/specifying-exec-params .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:DataflowCreateJavaJobOperator` :param jar: The reference to a self executing Dataflow jar (templated). :param job_name: The 'jobName' to use when executing the Dataflow job (templated). This ends up being set in the pipeline options, so any entry with key ``'jobName'`` in ``options`` will be overwritten. :param dataflow_default_options: Map of default job options. :param options: Map of job specific options.The key must be a dictionary. The value can contain different types: * If the value is None, the single option - ``--key`` (without value) will be added. * If the value is False, this option will be skipped * If the value is True, the single option - ``--key`` (without value) will be added. * If the value is list, the many options will be added for each key. If the value is ``['A', 'B']`` and the key is ``key`` then the ``--key=A --key=B`` options will be left * Other value types will be replaced with the Python textual representation. When defining labels (``labels`` option), you can also provide a dictionary. :param project_id: Optional, the Google Cloud project ID in which to start a job. If set to None or missing, the default project_id from the Google Cloud connection is used. :param location: Job location. :param gcp_conn_id: The connection ID to use connecting to Google Cloud. :param delegate_to: The account to impersonate using domain-wide delegation of authority, if any. For this to work, the service account making the request must have domain-wide delegation enabled. :param poll_sleep: The time in seconds to sleep between polling Google Cloud Platform for the dataflow job status while the job is in the JOB_STATE_RUNNING state. :param job_class: The name of the dataflow job class to be executed, it is often not the main class configured in the dataflow jar file. :param multiple_jobs: If pipeline creates multiple jobs then monitor all jobs :param check_if_running: before running job, validate that a previous run is not in process if job is running finish with nothing, WaitForRun= wait until job finished and the run job) ``jar``, ``options``, and ``job_name`` are templated so you can use variables in them. :param cancel_timeout: How long (in seconds) operator should wait for the pipeline to be successfully cancelled when task is being killed. :param wait_until_finished: (Optional) If True, wait for the end of pipeline execution before exiting. If False, only submits job. If None, default behavior. The default behavior depends on the type of pipeline: * for the streaming pipeline, wait for jobs to start, * for the batch pipeline, wait for the jobs to complete. .. warning:: You cannot call ``PipelineResult.wait_until_finish`` method in your pipeline code for the operator to work properly. i. e. you must use asynchronous execution. Otherwise, your pipeline will always wait until finished. For more information, look at: `Asynchronous execution <https://cloud.google.com/dataflow/docs/guides/specifying-exec-params#python_10>`__ The process of starting the Dataflow job in Airflow consists of two steps: * running a subprocess and reading the stderr/stderr log for the job id. * loop waiting for the end of the job ID from the previous step. This loop checks the status of the job. Step two is started just after step one has finished, so if you have wait_until_finished in your pipeline code, step two will not start until the process stops. When this process stops, steps two will run, but it will only execute one iteration as the job will be in a terminal state. If you in your pipeline do not call the wait_for_pipeline method but pass wait_until_finish=True to the operator, the second loop will wait for the job's terminal state. If you in your pipeline do not call the wait_for_pipeline method, and pass wait_until_finish=False to the operator, the second loop will check once is job not in terminal state and exit the loop. Note that both ``dataflow_default_options`` and ``options`` will be merged to specify pipeline execution parameter, and ``dataflow_default_options`` is expected to save high-level options, for instances, project and zone information, which apply to all dataflow operators in the DAG. It's a good practice to define dataflow_* parameters in the default_args of the dag like the project, zone and staging location. .. code-block:: python default_args = { "dataflow_default_options": { "zone": "europe-west1-d", "stagingLocation": "gs://my-staging-bucket/staging/", } } You need to pass the path to your dataflow as a file reference with the ``jar`` parameter, the jar needs to be a self executing jar (see documentation here: https://beam.apache.org/documentation/runners/dataflow/#self-executing-jar). Use ``options`` to pass on options to your job. .. code-block:: python t1 = DataflowCreateJavaJobOperator( task_id="dataflow_example", jar="{{var.value.gcp_dataflow_base}}pipeline/build/libs/pipeline-example-1.0.jar", options={ "autoscalingAlgorithm": "BASIC", "maxNumWorkers": "50", "start": "{{ds}}", "partitionType": "DAY", "labels": {"foo": "bar"}, }, gcp_conn_id="airflow-conn-id", dag=my - dag, ) """ template_fields: Sequence[str] = ("options", "jar", "job_name") ui_color = "#0273d4" def __init__( self, *, jar: str, job_name: str = "{{task.task_id}}", dataflow_default_options: Optional[dict] = None, options: Optional[dict] = None, project_id: Optional[str] = None, location: str = DEFAULT_DATAFLOW_LOCATION, gcp_conn_id: str = "google_cloud_default", delegate_to: Optional[str] = None, poll_sleep: int = 10, job_class: Optional[str] = None, check_if_running: CheckJobRunning = CheckJobRunning.WaitForRun, multiple_jobs: bool = False, cancel_timeout: Optional[int] = 10 * 60, wait_until_finished: Optional[bool] = None, **kwargs, ) -> None: # TODO: Remove one day warnings.warn( f"The `{self.__class__.__name__}` operator is deprecated, " f"please use `providers.apache.beam.operators.beam.BeamRunJavaPipelineOperator` instead.", DeprecationWarning, stacklevel=2, ) super().__init__(**kwargs) dataflow_default_options = dataflow_default_options or {} options = options or {} options.setdefault("labels", {}).update({ "airflow-version": "v" + version.replace(".", "-").replace("+", "-") }) self.project_id = project_id self.location = location self.gcp_conn_id = gcp_conn_id self.delegate_to = delegate_to self.jar = jar self.multiple_jobs = multiple_jobs self.job_name = job_name self.dataflow_default_options = dataflow_default_options self.options = options self.poll_sleep = poll_sleep self.job_class = job_class self.check_if_running = check_if_running self.cancel_timeout = cancel_timeout self.wait_until_finished = wait_until_finished self.job_id = None self.beam_hook: Optional[BeamHook] = None self.dataflow_hook: Optional[DataflowHook] = None def execute(self, context: 'Context'): """Execute the Apache Beam Pipeline.""" self.beam_hook = BeamHook(runner=BeamRunnerType.DataflowRunner) self.dataflow_hook = DataflowHook( gcp_conn_id=self.gcp_conn_id, delegate_to=self.delegate_to, poll_sleep=self.poll_sleep, cancel_timeout=self.cancel_timeout, wait_until_finished=self.wait_until_finished, ) job_name = self.dataflow_hook.build_dataflow_job_name( job_name=self.job_name) pipeline_options = copy.deepcopy(self.dataflow_default_options) pipeline_options["jobName"] = self.job_name pipeline_options[ "project"] = self.project_id or self.dataflow_hook.project_id pipeline_options["region"] = self.location pipeline_options.update(self.options) pipeline_options.setdefault("labels", {}).update({ "airflow-version": "v" + version.replace(".", "-").replace("+", "-") }) pipeline_options.update(self.options) def set_current_job_id(job_id): self.job_id = job_id process_line_callback = process_line_and_extract_dataflow_job_id_callback( on_new_job_id_callback=set_current_job_id) with ExitStack() as exit_stack: if self.jar.lower().startswith("gs://"): gcs_hook = GCSHook(self.gcp_conn_id, self.delegate_to) tmp_gcs_file = exit_stack.enter_context( gcs_hook.provide_file(object_url=self.jar)) self.jar = tmp_gcs_file.name is_running = False if self.check_if_running != CheckJobRunning.IgnoreJob: is_running = self.dataflow_hook.is_job_dataflow_running( name=self.job_name, variables=pipeline_options, ) while is_running and self.check_if_running == CheckJobRunning.WaitForRun: is_running = self.dataflow_hook.is_job_dataflow_running( name=self.job_name, variables=pipeline_options, ) if not is_running: pipeline_options["jobName"] = job_name with self.dataflow_hook.provide_authorized_gcloud(): self.beam_hook.start_java_pipeline( variables=pipeline_options, jar=self.jar, job_class=self.job_class, process_line_callback=process_line_callback, ) self.dataflow_hook.wait_for_done( job_name=job_name, location=self.location, job_id=self.job_id, multiple_jobs=self.multiple_jobs, ) return {"job_id": self.job_id} def on_kill(self) -> None: self.log.info("On kill.") if self.job_id: self.dataflow_hook.cancel_job(job_id=self.job_id, project_id=self.project_id or self.dataflow_hook.project_id)