You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@airflow.apache.org by GitBox <gi...@apache.org> on 2021/01/20 19:17:59 UTC
[GitHub] [airflow] aaltay commented on a change in pull request #12814: Add Apache Beam operators
aaltay commented on a change in pull request #12814:
URL: https://github.com/apache/airflow/pull/12814#discussion_r561159658
##########
File path: airflow/providers/apache/beam/BACKPORT_PROVIDER_README.md
##########
@@ -0,0 +1,101 @@
+<!--
+ Licensed to the Apache Software Foundation (ASF) under one
+ or more contributor license agreements. See the NOTICE file
+ distributed with this work for additional information
+ regarding copyright ownership. The ASF licenses this file
+ to you under the Apache License, Version 2.0 (the
+ "License"); you may not use this file except in compliance
+ with the License. You may obtain a copy of the License at
+
+ http://www.apache.org/licenses/LICENSE-2.0
+
+ Unless required by applicable law or agreed to in writing,
+ software distributed under the License is distributed on an
+ "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ KIND, either express or implied. See the License for the
+ specific language governing permissions and limitations
+ under the License.
+ -->
+
+
+# Package apache-airflow-backport-providers-apache-beam
+
+Release:
+
+**Table of contents**
+
+- [Backport package](#backport-package)
+- [Installation](#installation)
+- [PIP requirements](#pip-requirements)
+- [Cross provider package dependencies](#cross-provider-package-dependencies)
+- [Provider class summary](#provider-classes-summary)
+ - [Operators](#operators)
+ - [Moved operators](#moved-operators)
+ - [Transfer operators](#transfer-operators)
+ - [Moved transfer operators](#moved-transfer-operators)
+ - [Hooks](#hooks)
+ - [Moved hooks](#moved-hooks)
+- [Releases](#releases)
+ - [Release](#release)
+
+## Backport package
+
+This is a backport providers package for `apache.beam` provider. All classes for this provider package
+are in `airflow.providers.apache.beam` python package.
+
+**Only Python 3.6+ is supported for this backport package.**
+
+While Airflow 1.10.* continues to support Python 2.7+ - you need to upgrade python to 3.6+ if you
+want to use this backport package.
+
+
+
+## Installation
+
+You can install this package on top of an existing airflow 1.10.* installation via
+`pip install apache-airflow-backport-providers-apache-beam`
+
+## PIP requirements
+
+
+## Cross provider package dependencies
+
+Those are dependencies that might be needed in order to use all the features of the package.
+You need to install the specified backport providers package in order to use them.
+
+You can install such cross-provider dependencies when installing from PyPI. For example:
+
+```bash
+pip install apache-airflow-backport-providers-apache-beam[google]
+```
+
+| Dependent package | Extra |
+|:--------------------------------------------------------------------------------------------------------------------|:-------|
+| [apache-airflow-backport-providers-google](https://github.com/apache/airflow/tree/master/airflow/providers/google/) | google |
+
+# Provider classes summary
+
+In Airflow 2.0, all operators, transfers, hooks, sensors, secrets for the `apache.beam` provider
+are in the `airflow.providers.apache.beam` package. You can read more about the naming conventions used
+in [Naming conventions for provider packages](https://github.com/apache/airflow/blob/master/CONTRIBUTING.rst#naming-conventions-for-provider-packages)
+
+
+## Operators
Review comment:
There is more that could be added here related to hooks, and operators maybe?
##########
File path: airflow/providers/apache/beam/hooks/beam.py
##########
@@ -0,0 +1,287 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains a Apache Beam Hook."""
+import json
+import select
+import shlex
+import subprocess
+import textwrap
+from tempfile import TemporaryDirectory
+from typing import Callable, List, Optional
+
+from airflow.exceptions import AirflowException
+from airflow.hooks.base_hook import BaseHook
+from airflow.utils.log.logging_mixin import LoggingMixin
+from airflow.utils.python_virtualenv import prepare_virtualenv
+
+
+class BeamRunnerType:
Review comment:
Why is this needed? Ideally any Beam supported runners should just work, and there is no need for this duplication.
##########
File path: airflow/providers/apache/beam/example_dags/example_beam.py
##########
@@ -0,0 +1,305 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+"""
+Example Airflow DAG for Apache Beam operators
+"""
+import os
+from urllib.parse import urlparse
+
+from airflow import models
+from airflow.providers.apache.beam.operators.beam import (
+ BeamRunJavaPipelineOperator,
+ BeamRunPythonPipelineOperator,
+)
+from airflow.providers.google.cloud.hooks.dataflow import DataflowJobStatus
+from airflow.providers.google.cloud.operators.dataflow import DataflowPythonConfiguration
+from airflow.providers.google.cloud.sensors.dataflow import DataflowJobStatusSensor
+from airflow.providers.google.cloud.transfers.gcs_to_local import GCSToLocalFilesystemOperator
+from airflow.utils.dates import days_ago
+
+GCP_PROJECT_ID = os.environ.get('GCP_PROJECT_ID', 'example-project')
+GCS_INPUT = os.environ.get('APACHE_BEAM_PYTHON', 'gs://apache-beam-samples/shakespeare/kinglear.txt')
+GCS_TMP = os.environ.get('APACHE_BEAM_GCS_TMP', 'gs://test-dataflow-example/temp/')
+GCS_STAGING = os.environ.get('APACHE_BEAM_GCS_STAGING', 'gs://test-dataflow-example/staging/')
+GCS_OUTPUT = os.environ.get('APACHE_BEAM_GCS_OUTPUT', 'gs://test-dataflow-example/output')
+GCS_PYTHON = os.environ.get('APACHE_BEAM_PYTHON', 'gs://test-dataflow-example/wordcount_debugging.py')
+GCS_PYTHON_DATAFLOW_ASYNC = os.environ.get(
+ 'APACHE_BEAM_PYTHON_DATAFLOW_ASYNC', 'gs://test-dataflow-example/wordcount_debugging.py'
+)
+
+GCS_JAR_DIRECT_RUNNER = os.environ.get(
+ 'APACHE_BEAM_DIRECT_RUNNER_JAR',
+ 'gs://test-dataflow-example/tests/dataflow-templates-bundled-java=11-beam-v2.25.0-DirectRunner.jar',
+)
+GCS_JAR_DATAFLOW_RUNNER = os.environ.get(
+ 'APACHE_BEAM_DATAFLOW_RUNNER_JAR', 'gs://test-dataflow-example/word-count-beam-bundled-0.1.jar'
+)
+GCS_JAR_SPARK_RUNNER = os.environ.get(
+ 'APACHE_BEAM_SPARK_RUNNER_JAR',
+ 'gs://test-dataflow-example/tests/dataflow-templates-bundled-java=11-beam-v2.25.0-SparkRunner.jar',
+)
+GCS_JAR_FLINK_RUNNER = os.environ.get(
+ 'APACHE_BEAM_FLINK_RUNNER_JAR',
+ 'gs://test-dataflow-example/tests/dataflow-templates-bundled-java=11-beam-v2.25.0-FlinkRunner.jar',
+)
+
+GCS_JAR_DIRECT_RUNNER_PARTS = urlparse(GCS_JAR_DIRECT_RUNNER)
+GCS_JAR_DIRECT_RUNNER_BUCKET_NAME = GCS_JAR_DIRECT_RUNNER_PARTS.netloc
+GCS_JAR_DIRECT_RUNNER_OBJECT_NAME = GCS_JAR_DIRECT_RUNNER_PARTS.path[1:]
+GCS_JAR_DATAFLOW_RUNNER_PARTS = urlparse(GCS_JAR_DATAFLOW_RUNNER)
+GCS_JAR_DATAFLOW_RUNNER_BUCKET_NAME = GCS_JAR_DATAFLOW_RUNNER_PARTS.netloc
+GCS_JAR_DATAFLOW_RUNNER_OBJECT_NAME = GCS_JAR_DATAFLOW_RUNNER_PARTS.path[1:]
+GCS_JAR_SPARK_RUNNER_PARTS = urlparse(GCS_JAR_SPARK_RUNNER)
+GCS_JAR_SPARK_RUNNER_BUCKET_NAME = GCS_JAR_SPARK_RUNNER_PARTS.netloc
+GCS_JAR_SPARK_RUNNER_OBJECT_NAME = GCS_JAR_SPARK_RUNNER_PARTS.path[1:]
+GCS_JAR_FLINK_RUNNER_PARTS = urlparse(GCS_JAR_FLINK_RUNNER)
+GCS_JAR_FLINK_RUNNER_BUCKET_NAME = GCS_JAR_FLINK_RUNNER_PARTS.netloc
+GCS_JAR_FLINK_RUNNER_OBJECT_NAME = GCS_JAR_FLINK_RUNNER_PARTS.path[1:]
+
+
+default_args = {
+ 'default_pipeline_options': {
+ 'output': '/tmp/example_beam',
+ },
+ "trigger_rule": "all_done",
+}
+
+
+with models.DAG(
+ "example_beam_native_java_direct_runner",
+ schedule_interval=None, # Override to match your needs
+ start_date=days_ago(1),
+ tags=['example'],
+) as dag_native_java_direct_runner:
+
+ jar_to_local_direct_runner = GCSToLocalFilesystemOperator(
+ task_id="jar_to_local_direct_runner",
+ bucket=GCS_JAR_DIRECT_RUNNER_BUCKET_NAME,
+ object_name=GCS_JAR_DIRECT_RUNNER_OBJECT_NAME,
+ filename="/tmp/beam_wordcount_direct_runner_{{ ds_nodash }}.jar",
+ )
+
+ start_java_pipeline_direct_runner = BeamRunJavaPipelineOperator(
+ task_id="start_java_pipeline_direct_runner",
+ runner="DirectRunner",
Review comment:
Is DirectRunner the default option? (If yes, you can omit this arg.)
##########
File path: airflow/providers/apache/beam/operators/beam.py
##########
@@ -0,0 +1,469 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains Apache Beam operators."""
+import re
+from contextlib import ExitStack
+from typing import Callable, List, Optional, Union
+
+from airflow.models import BaseOperator
+from airflow.providers.apache.beam.hooks.beam import BeamHook, BeamRunnerType
+from airflow.providers.google.cloud.hooks.dataflow import (
+ DataflowHook,
+ process_line_and_extract_dataflow_job_id_callback,
+)
+from airflow.providers.google.cloud.hooks.gcs import GCSHook
+from airflow.providers.google.cloud.operators.dataflow import (
+ CheckJobRunning,
+ DataflowJavaConfiguration,
+ DataflowPythonConfiguration,
+)
+from airflow.utils.decorators import apply_defaults
+from airflow.version import version
+
+
+class BeamRunPythonPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Python. Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level options, for instances, project and zone information, which
+ apply to all beam operators in the DAG.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ with models.DAG(
+ "example_beam_native_python",
+ default_args=default_args,
+ start_date=days_ago(1),
+ schedule_interval=None,
+ tags=['example'],
+ ) as dag_native_python:
+
+ start_python_job_local_direct_runner = BeamRunPythonPipelineOperator(
+ task_id="start_python_job_local_direct_runner",
+ runner="DirectRunner",
+ py_file='apache_beam.examples.wordcount',
+ py_options=['-m'],
+ py_requirements=['apache-beam[gcp]==2.21.0'],
+ py_interpreter='python3',
+ py_system_site_packages=False,
+ )
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param py_file: Reference to the python Apache Beam pipeline file.py, e.g.,
+ /some/local/file/path/to/your/python/pipeline/file. (templated)
+ :type py_file: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ Other possible options: DataflowRunner, SparkRunner, FlinkRunner.
+ See: :class:`~providers.apache.beam.hooks.beam.BeamRunnerType`
+ See: https://beam.apache.org/documentation/runners/capability-matrix/
+
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreatePythonJobOperator`
+ :type runner: str
+ :param py_options: Additional python options, e.g., ["-m", "-v"].
+ :type py_options: list[str]
+ :param default_pipeline_options: Map of default pipeline options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of pipeline 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.
+ :type pipeline_options: dict
+ :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
+ :type py_interpreter: str
+ :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.
+ :type py_requirements: List[str]
+ :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: Optional.
+ The connection ID to use connecting to Google Cloud Storage if pyfile is on GCS.
+ :type gcp_conn_id: str
+ :param delegate_to: Optional.
+ 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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowPythonConfiguration]
+ """
+
+ template_fields = ["py_file", "runner", "pipeline_options", "default_pipeline_options", "dataflow_config"]
+ template_fields_renderers = {'dataflow_config': 'json'}
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ py_file: str,
+ runner: str = "DirectRunner",
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_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,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowPythonConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.py_file = py_file
+ self.runner = runner
+ self.py_options = py_options or []
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.pipeline_options.setdefault("labels", {}).update(
+ {"airflow-version": "v" + version.replace(".", "-").replace("+", "-")}
Review comment:
Does the operator always have the same version as airflow? Would it make sense to include what is the operator version that is used?
##########
File path: airflow/providers/apache/beam/operators/beam.py
##########
@@ -0,0 +1,469 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains Apache Beam operators."""
+import re
+from contextlib import ExitStack
+from typing import Callable, List, Optional, Union
+
+from airflow.models import BaseOperator
+from airflow.providers.apache.beam.hooks.beam import BeamHook, BeamRunnerType
+from airflow.providers.google.cloud.hooks.dataflow import (
+ DataflowHook,
+ process_line_and_extract_dataflow_job_id_callback,
+)
+from airflow.providers.google.cloud.hooks.gcs import GCSHook
+from airflow.providers.google.cloud.operators.dataflow import (
+ CheckJobRunning,
+ DataflowJavaConfiguration,
+ DataflowPythonConfiguration,
+)
+from airflow.utils.decorators import apply_defaults
+from airflow.version import version
+
+
+class BeamRunPythonPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Python. Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level options, for instances, project and zone information, which
+ apply to all beam operators in the DAG.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ with models.DAG(
+ "example_beam_native_python",
+ default_args=default_args,
+ start_date=days_ago(1),
+ schedule_interval=None,
+ tags=['example'],
+ ) as dag_native_python:
+
+ start_python_job_local_direct_runner = BeamRunPythonPipelineOperator(
+ task_id="start_python_job_local_direct_runner",
+ runner="DirectRunner",
+ py_file='apache_beam.examples.wordcount',
+ py_options=['-m'],
+ py_requirements=['apache-beam[gcp]==2.21.0'],
+ py_interpreter='python3',
+ py_system_site_packages=False,
+ )
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param py_file: Reference to the python Apache Beam pipeline file.py, e.g.,
+ /some/local/file/path/to/your/python/pipeline/file. (templated)
+ :type py_file: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ Other possible options: DataflowRunner, SparkRunner, FlinkRunner.
+ See: :class:`~providers.apache.beam.hooks.beam.BeamRunnerType`
+ See: https://beam.apache.org/documentation/runners/capability-matrix/
+
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreatePythonJobOperator`
+ :type runner: str
+ :param py_options: Additional python options, e.g., ["-m", "-v"].
+ :type py_options: list[str]
+ :param default_pipeline_options: Map of default pipeline options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of pipeline 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.
+ :type pipeline_options: dict
+ :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
+ :type py_interpreter: str
+ :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.
+ :type py_requirements: List[str]
+ :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: Optional.
+ The connection ID to use connecting to Google Cloud Storage if pyfile is on GCS.
+ :type gcp_conn_id: str
+ :param delegate_to: Optional.
+ 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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowPythonConfiguration]
+ """
+
+ template_fields = ["py_file", "runner", "pipeline_options", "default_pipeline_options", "dataflow_config"]
+ template_fields_renderers = {'dataflow_config': 'json'}
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ py_file: str,
+ runner: str = "DirectRunner",
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_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,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowPythonConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.py_file = py_file
+ self.runner = runner
+ self.py_options = py_options or []
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.pipeline_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.gcp_conn_id = gcp_conn_id
+ self.delegate_to = delegate_to
+ self.dataflow_config = dataflow_config or {}
+ self.beam_hook: Optional[BeamHook] = None
+ self.dataflow_hook: Optional[DataflowHook] = None
+ self.dataflow_job_id: Optional[str] = None
+ self._dataflow_job_name: Optional[str] = None
Review comment:
_dataflow_job_name - I think this does not need to be a member. It seems to be local only to execute().
##########
File path: airflow/providers/apache/beam/operators/beam.py
##########
@@ -0,0 +1,469 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains Apache Beam operators."""
+import re
+from contextlib import ExitStack
+from typing import Callable, List, Optional, Union
+
+from airflow.models import BaseOperator
+from airflow.providers.apache.beam.hooks.beam import BeamHook, BeamRunnerType
+from airflow.providers.google.cloud.hooks.dataflow import (
+ DataflowHook,
+ process_line_and_extract_dataflow_job_id_callback,
+)
+from airflow.providers.google.cloud.hooks.gcs import GCSHook
+from airflow.providers.google.cloud.operators.dataflow import (
+ CheckJobRunning,
+ DataflowJavaConfiguration,
+ DataflowPythonConfiguration,
+)
+from airflow.utils.decorators import apply_defaults
+from airflow.version import version
+
+
+class BeamRunPythonPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Python. Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level options, for instances, project and zone information, which
+ apply to all beam operators in the DAG.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ with models.DAG(
+ "example_beam_native_python",
+ default_args=default_args,
+ start_date=days_ago(1),
+ schedule_interval=None,
+ tags=['example'],
+ ) as dag_native_python:
+
+ start_python_job_local_direct_runner = BeamRunPythonPipelineOperator(
+ task_id="start_python_job_local_direct_runner",
+ runner="DirectRunner",
+ py_file='apache_beam.examples.wordcount',
+ py_options=['-m'],
+ py_requirements=['apache-beam[gcp]==2.21.0'],
+ py_interpreter='python3',
+ py_system_site_packages=False,
+ )
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param py_file: Reference to the python Apache Beam pipeline file.py, e.g.,
+ /some/local/file/path/to/your/python/pipeline/file. (templated)
+ :type py_file: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ Other possible options: DataflowRunner, SparkRunner, FlinkRunner.
+ See: :class:`~providers.apache.beam.hooks.beam.BeamRunnerType`
+ See: https://beam.apache.org/documentation/runners/capability-matrix/
+
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreatePythonJobOperator`
+ :type runner: str
+ :param py_options: Additional python options, e.g., ["-m", "-v"].
+ :type py_options: list[str]
+ :param default_pipeline_options: Map of default pipeline options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of pipeline 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.
+ :type pipeline_options: dict
+ :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
+ :type py_interpreter: str
+ :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.
+ :type py_requirements: List[str]
+ :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: Optional.
+ The connection ID to use connecting to Google Cloud Storage if pyfile is on GCS.
+ :type gcp_conn_id: str
+ :param delegate_to: Optional.
+ 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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowPythonConfiguration]
+ """
+
+ template_fields = ["py_file", "runner", "pipeline_options", "default_pipeline_options", "dataflow_config"]
+ template_fields_renderers = {'dataflow_config': 'json'}
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ py_file: str,
+ runner: str = "DirectRunner",
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_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,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowPythonConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.py_file = py_file
+ self.runner = runner
+ self.py_options = py_options or []
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.pipeline_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.gcp_conn_id = gcp_conn_id
+ self.delegate_to = delegate_to
+ self.dataflow_config = dataflow_config or {}
+ self.beam_hook: Optional[BeamHook] = None
+ self.dataflow_hook: Optional[DataflowHook] = None
+ self.dataflow_job_id: Optional[str] = None
+ self._dataflow_job_name: Optional[str] = None
+
+ def execute(self, context):
+ """Execute the Apache Beam Pipeline."""
+ self.beam_hook = BeamHook(runner=self.runner)
+ pipeline_options = self.default_pipeline_options.copy()
+ process_line_callback: Optional[Callable] = None
+
+ if isinstance(self.dataflow_config, dict):
+ self.dataflow_config = DataflowPythonConfiguration(**self.dataflow_config)
Review comment:
DataflowPythonConfiguration is getting a bit strange in Beam operator. I wonder if it can be merged into regular options?
Another question is there a reason for DataflowPythonConfiguration / and its java equivalent. Presumably they are similar and this is more like "DataflowConfiguration"
##########
File path: airflow/providers/google/cloud/hooks/dataflow.py
##########
@@ -50,6 +47,35 @@
T = TypeVar("T", bound=Callable) # pylint: disable=invalid-name
+def process_line_and_extract_dataflow_job_id_callback(
+ on_new_job_id_callback: Optional[Callable[[str], None]]
+) -> Callable[[str], None]:
+ """
+ Returns callback which triggers function passed as `on_new_job_id_callback` when Dataflow job_id is found.
+ To be used for `process_line_callback` in
+ :py:class:`~airflow.providers.apache.beam.hooks.beam.BeamCommandRunner`
+
+ :param on_new_job_id_callback: Callback called when the job ID is known
+ :type on_new_job_id_callback: callback
+ """
+
+ def _process_line_and_extract_job_id(
+ line: str,
+ # on_new_job_id_callback: Optional[Callable[[str], None]]
+ ) -> None:
+ # Job id info: https://goo.gl/SE29y9.
+ matched_job = JOB_ID_PATTERN.search(line)
+ if matched_job:
+ job_id = matched_job.group("job_id_java") or matched_job.group("job_id_python")
Review comment:
What is the difference between job_id_java and job_id_python?
##########
File path: airflow/providers/google/cloud/operators/dataflow.py
##########
@@ -43,6 +48,214 @@ class CheckJobRunning(Enum):
WaitForRun = 3
+class DataflowConfiguration(metaclass=ABCMeta):
+ """Abstract class for Dataflow configuration to be passed to Beam operators"""
+
+ template_fields = ["job_name", "location"]
+
+ def __init__(
+ self,
+ *,
+ job_name: Optional[str] = "{{task.task_id}}",
+ append_job_name: bool = True,
+ project_id: Optional[str] = None,
+ location: Optional[str] = DEFAULT_DATAFLOW_LOCATION,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ poll_sleep: int = 10,
+ impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
+ drain_pipeline: bool = False,
+ cancel_timeout: Optional[int] = 5 * 60,
+ wait_until_finished: Optional[bool] = None,
+ ) -> None:
+ self.job_name = job_name
+ self.append_job_name = append_job_name
+ 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.impersonation_chain = impersonation_chain
+ self.drain_pipeline = drain_pipeline
+ self.cancel_timeout = cancel_timeout
+ self.wait_until_finished = wait_until_finished
+
+
+class DataflowPythonConfiguration(DataflowConfiguration):
Review comment:
I wonder if you can remove DataflowPythonConfiguration and java version, in favor of only keeping the base version. (E.g. we could avoid adding another one when Go is available.)
##########
File path: airflow/providers/google/cloud/hooks/dataflow.py
##########
@@ -680,22 +582,36 @@ def start_java_dataflow(
:param location: Job location.
:type location: str
"""
- name = self._build_dataflow_job_name(job_name, append_job_name)
+ warnings.warn(
+ """"This method is deprecated.
Review comment:
Are the new versions fully backward compatible with these ones?
##########
File path: airflow/providers/apache/beam/operators/beam.py
##########
@@ -0,0 +1,469 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains Apache Beam operators."""
+import re
+from contextlib import ExitStack
+from typing import Callable, List, Optional, Union
+
+from airflow.models import BaseOperator
+from airflow.providers.apache.beam.hooks.beam import BeamHook, BeamRunnerType
+from airflow.providers.google.cloud.hooks.dataflow import (
+ DataflowHook,
+ process_line_and_extract_dataflow_job_id_callback,
+)
+from airflow.providers.google.cloud.hooks.gcs import GCSHook
+from airflow.providers.google.cloud.operators.dataflow import (
+ CheckJobRunning,
+ DataflowJavaConfiguration,
+ DataflowPythonConfiguration,
+)
+from airflow.utils.decorators import apply_defaults
+from airflow.version import version
+
+
+class BeamRunPythonPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Python. Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level options, for instances, project and zone information, which
+ apply to all beam operators in the DAG.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ with models.DAG(
+ "example_beam_native_python",
+ default_args=default_args,
+ start_date=days_ago(1),
+ schedule_interval=None,
+ tags=['example'],
+ ) as dag_native_python:
+
+ start_python_job_local_direct_runner = BeamRunPythonPipelineOperator(
+ task_id="start_python_job_local_direct_runner",
+ runner="DirectRunner",
+ py_file='apache_beam.examples.wordcount',
+ py_options=['-m'],
+ py_requirements=['apache-beam[gcp]==2.21.0'],
+ py_interpreter='python3',
+ py_system_site_packages=False,
+ )
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param py_file: Reference to the python Apache Beam pipeline file.py, e.g.,
+ /some/local/file/path/to/your/python/pipeline/file. (templated)
+ :type py_file: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ Other possible options: DataflowRunner, SparkRunner, FlinkRunner.
+ See: :class:`~providers.apache.beam.hooks.beam.BeamRunnerType`
+ See: https://beam.apache.org/documentation/runners/capability-matrix/
+
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreatePythonJobOperator`
+ :type runner: str
+ :param py_options: Additional python options, e.g., ["-m", "-v"].
+ :type py_options: list[str]
+ :param default_pipeline_options: Map of default pipeline options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of pipeline 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.
+ :type pipeline_options: dict
+ :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
+ :type py_interpreter: str
+ :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.
+ :type py_requirements: List[str]
+ :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: Optional.
+ The connection ID to use connecting to Google Cloud Storage if pyfile is on GCS.
+ :type gcp_conn_id: str
+ :param delegate_to: Optional.
+ 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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowPythonConfiguration]
+ """
+
+ template_fields = ["py_file", "runner", "pipeline_options", "default_pipeline_options", "dataflow_config"]
+ template_fields_renderers = {'dataflow_config': 'json'}
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ py_file: str,
+ runner: str = "DirectRunner",
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_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,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowPythonConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.py_file = py_file
+ self.runner = runner
+ self.py_options = py_options or []
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.pipeline_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.gcp_conn_id = gcp_conn_id
+ self.delegate_to = delegate_to
+ self.dataflow_config = dataflow_config or {}
+ self.beam_hook: Optional[BeamHook] = None
+ self.dataflow_hook: Optional[DataflowHook] = None
+ self.dataflow_job_id: Optional[str] = None
+ self._dataflow_job_name: Optional[str] = None
+
+ def execute(self, context):
+ """Execute the Apache Beam Pipeline."""
+ self.beam_hook = BeamHook(runner=self.runner)
+ pipeline_options = self.default_pipeline_options.copy()
+ process_line_callback: Optional[Callable] = None
+
+ if isinstance(self.dataflow_config, dict):
+ self.dataflow_config = DataflowPythonConfiguration(**self.dataflow_config)
+
+ if self.runner.lower() == BeamRunnerType.DataflowRunner.lower():
+ self.dataflow_hook = DataflowHook(
+ gcp_conn_id=self.dataflow_config.gcp_conn_id or self.gcp_conn_id,
+ delegate_to=self.dataflow_config.delegate_to or self.delegate_to,
+ poll_sleep=self.dataflow_config.poll_sleep,
+ impersonation_chain=self.dataflow_config.impersonation_chain,
+ drain_pipeline=self.dataflow_config.drain_pipeline,
+ cancel_timeout=self.dataflow_config.cancel_timeout,
+ wait_until_finished=self.dataflow_config.wait_until_finished,
+ )
+ self.dataflow_config.project_id = self.dataflow_config.project_id or self.dataflow_hook.project_id
+
+ self._dataflow_job_name = DataflowHook.build_dataflow_job_name(
+ self.dataflow_config.job_name, self.dataflow_config.append_job_name
+ )
+ pipeline_options["job_name"] = self._dataflow_job_name
+ pipeline_options["project"] = self.dataflow_config.project_id
+ pipeline_options["region"] = self.dataflow_config.location
+ pipeline_options.setdefault("labels", {}).update(
+ {"airflow-version": "v" + version.replace(".", "-").replace("+", "-")}
+ )
+
+ def set_current_dataflow_job_id(job_id):
+ self.dataflow_job_id = job_id
+
+ process_line_callback = process_line_and_extract_dataflow_job_id_callback(
+ on_new_job_id_callback=set_current_dataflow_job_id
+ )
+
+ pipeline_options.update(self.pipeline_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}
+
+ 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( # pylint: disable=no-member
+ gcs_hook.provide_file(object_url=self.py_file)
+ )
+ self.py_file = tmp_gcs_file.name
+
+ 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,
+ )
+
+ if self.runner.lower() == BeamRunnerType.DataflowRunner.lower():
+ self.dataflow_hook.wait_for_done( # pylint: disable=no-value-for-parameter
+ job_name=self._dataflow_job_name,
+ location=self.dataflow_config.location,
+ job_id=self.dataflow_job_id,
+ multiple_jobs=False,
+ )
+
+ return {"dataflow_job_id": self.dataflow_job_id}
+
+ def on_kill(self) -> None:
+ self.log.info("On kill.")
+ if self.dataflow_hook and self.dataflow_job_id:
+ self.dataflow_hook.cancel_job(
+ job_id=self.dataflow_job_id,
+ project_id=self.dataflow_config.project_id,
+ )
+
+
+# pylint: disable=too-many-instance-attributes
+class BeamRunJavaPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Java.
+
+ Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level pipeline_options, for instances, project and zone information, which
+ apply to all Apache Beam operators in the DAG.
+
+ It's a good practice to define parameters in the default_args of the dag
+ like the project, zone and staging location.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ You need to pass the path to your jar file 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 ``pipeline_options`` to pass on pipeline_options to your job.
+
+ .. code-block:: python
+
+ t1 = BeamRunJavaPipelineOperator(
+ task_id='start_java_job_spark_runner',
+ jar='{{var.value.spark_runner_jar_base}}pipeline/build/libs/pipeline-example-1.0.jar',
+ pipeline_options={
+ 'output': '/tmp/start_java_job_spark_runner',
+ 'inputFile': 'gs://apache-beam-samples/shakespeare/kinglear.txt,
+ },
+ dag=my-dag)
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param jar: The reference to a self executing Apache Beam jar (templated).
+ :type jar: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ See:
+ https://beam.apache.org/documentation/runners/capability-matrix/
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreateJavaJobOperator`
+ :type runner: str
+ :param job_class: The name of the Apache Beam pipeline class to be executed, it
+ is often not the main class configured in the pipeline jar file.
+ :type job_class: str
+ :param default_pipeline_options: Map of default job pipeline_options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of job specific pipeline_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 pipeline_options will be added for each key.
+ If the value is ``['A', 'B']`` and the key is ``key`` then the ``--key=A --key-B`` pipeline_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.
+ :type pipeline_options: dict
+ :param gcp_conn_id: The connection ID to use connecting to Google Cloud Storage if jar is on GCS
+ :type gcp_conn_id: str
+ :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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowJavaConfiguration]
+ """
+
+ template_fields = [
+ "jar",
+ "runner",
+ "job_class",
+ "pipeline_options",
+ "default_pipeline_options",
+ "dataflow_config",
+ ]
+ template_fields_renderers = {'dataflow_config': 'json'}
+ ui_color = "#0273d4"
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ jar: str,
+ runner: str = "DirectRunner",
+ job_class: Optional[str] = None,
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_options: Optional[dict] = None,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowJavaConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.jar = jar
+ self.runner = runner
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.job_class = job_class
+ self.dataflow_config = dataflow_config or {}
+ self.gcp_conn_id = gcp_conn_id
+ self.delegate_to = delegate_to
+ self.dataflow_job_id = None
+ self.dataflow_hook: Optional[DataflowHook] = None
+ self.beam_hook: Optional[BeamHook] = None
+ self._dataflow_job_name: Optional[str] = None
+
+ def execute(self, context):
+ """Execute the Apache Beam Pipeline."""
+ self.beam_hook = BeamHook(runner=self.runner)
+ pipeline_options = self.default_pipeline_options.copy()
+ process_line_callback: Optional[Callable] = None
+
+ if isinstance(self.dataflow_config, dict):
+ self.dataflow_config = DataflowJavaConfiguration(**self.dataflow_config)
Review comment:
what happens if dataflow config is not none, but the runner is something other than dataflow runner. Should it be an error/ warning?
##########
File path: airflow/providers/google/cloud/operators/dataflow.py
##########
@@ -43,6 +48,214 @@ class CheckJobRunning(Enum):
WaitForRun = 3
+class DataflowConfiguration(metaclass=ABCMeta):
+ """Abstract class for Dataflow configuration to be passed to Beam operators"""
+
+ template_fields = ["job_name", "location"]
+
+ def __init__(
+ self,
+ *,
+ job_name: Optional[str] = "{{task.task_id}}",
+ append_job_name: bool = True,
+ project_id: Optional[str] = None,
+ location: Optional[str] = DEFAULT_DATAFLOW_LOCATION,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ poll_sleep: int = 10,
+ impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
+ drain_pipeline: bool = False,
+ cancel_timeout: Optional[int] = 5 * 60,
+ wait_until_finished: Optional[bool] = None,
+ ) -> None:
+ self.job_name = job_name
+ self.append_job_name = append_job_name
+ 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.impersonation_chain = impersonation_chain
+ self.drain_pipeline = drain_pipeline
+ self.cancel_timeout = cancel_timeout
+ self.wait_until_finished = wait_until_finished
+
+
+class DataflowPythonConfiguration(DataflowConfiguration):
+ """
+ Dataflow configuration that can be passed to
+ :py:class:`~airflow.providers.apache.beam.operators.beam.BeamRunPythonPipelineOperator`
+
+ :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'`` or ``'job_name'``in ``options`` will be overwritten.
+ :type job_name: str
+ :param append_job_name: True if unique suffix has to be appended to job name.
+ :type append_job_name: bool
+ :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.
+ :type project_id: str
+ :param location: Job location.
+ :type location: str
+ :param gcp_conn_id: The connection ID to use connecting to Google Cloud.
+ :type gcp_conn_id: str
+ :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.
+ :type delegate_to: str
+ :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.
+ :type poll_sleep: int
+ :param impersonation_chain: Optional service account to impersonate using short-term
+ credentials, or chained list of accounts required to get the access_token
+ of the last account in the list, which will be impersonated in the request.
+ If set as a string, the account must grant the originating account
+ the Service Account Token Creator IAM role.
+ If set as a sequence, the identities from the list must grant
+ Service Account Token Creator IAM role to the directly preceding identity, with first
+ account from the list granting this role to the originating account (templated).
+ :type impersonation_chain: Union[str, Sequence[str]]
+ :param drain_pipeline: Optional, set to True if want to stop streaming job by draining it
+ instead of canceling during during killing task instance. See:
+ https://cloud.google.com/dataflow/docs/guides/stopping-a-pipeline
+ :type drain_pipeline: bool
+ :param cancel_timeout: How long (in seconds) operator should wait for the pipeline to be
+ successfully cancelled when task is being killed.
+ :type cancel_timeout: Optional[int]
+ :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.
+ :type wait_until_finished: Optional[bool]
+ """
+
+
+class DataflowJavaConfiguration(DataflowConfiguration):
+ """
+ Dataflow configuration that can be passed to
+ :py:class:`~airflow.providers.apache.beam.operators.beam.BeamRunJavaPipelineOperator`
+
+ :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'`` or ``'job_name'``in ``options`` will be overwritten.
+ :type job_name: str
+ :param append_job_name: True if unique suffix has to be appended to job name.
+ :type append_job_name: bool
+ :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.
+ :type project_id: str
+ :param location: Job location.
+ :type location: str
+ :param gcp_conn_id: The connection ID to use connecting to Google Cloud.
+ :type gcp_conn_id: str
+ :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.
+ :type delegate_to: str
+ :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.
+ :type poll_sleep: int
+ :param impersonation_chain: Optional service account to impersonate using short-term
+ credentials, or chained list of accounts required to get the access_token
+ of the last account in the list, which will be impersonated in the request.
+ If set as a string, the account must grant the originating account
+ the Service Account Token Creator IAM role.
+ If set as a sequence, the identities from the list must grant
+ Service Account Token Creator IAM role to the directly preceding identity, with first
+ account from the list granting this role to the originating account (templated).
+ :type impersonation_chain: Union[str, Sequence[str]]
+ :param drain_pipeline: Optional, set to True if want to stop streaming job by draining it
+ instead of canceling during during killing task instance. See:
+ https://cloud.google.com/dataflow/docs/guides/stopping-a-pipeline
+ :type drain_pipeline: bool
+ :param cancel_timeout: How long (in seconds) operator should wait for the pipeline to be
+ successfully cancelled when task is being killed.
+ :type cancel_timeout: Optional[int]
+ :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.
+ :type wait_until_finished: Optional[bool]
+ :param multiple_jobs: If pipeline creates multiple jobs then monitor all jobs
+ :type multiple_jobs: boolean
+ :param check_if_running: before running job, validate that a previous run is not in process
+ :type check_if_running: CheckJobRunning(IgnoreJob = do not check if running, FinishIfRunning=
+ 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.
+ """
+
+ def __init__(
+ self,
+ *,
+ multiple_jobs: Optional[bool] = None,
+ check_if_running: CheckJobRunning = CheckJobRunning.WaitForRun,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+ self.multiple_jobs = multiple_jobs
+ self.check_if_running = check_if_running
+
+
# pylint: disable=too-many-instance-attributes
class DataflowCreateJavaJobOperator(BaseOperator):
Review comment:
Should this have a deprecation warning?
##########
File path: airflow/providers/apache/beam/operators/beam.py
##########
@@ -0,0 +1,469 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""This module contains Apache Beam operators."""
+import re
+from contextlib import ExitStack
+from typing import Callable, List, Optional, Union
+
+from airflow.models import BaseOperator
+from airflow.providers.apache.beam.hooks.beam import BeamHook, BeamRunnerType
+from airflow.providers.google.cloud.hooks.dataflow import (
+ DataflowHook,
+ process_line_and_extract_dataflow_job_id_callback,
+)
+from airflow.providers.google.cloud.hooks.gcs import GCSHook
+from airflow.providers.google.cloud.operators.dataflow import (
+ CheckJobRunning,
+ DataflowJavaConfiguration,
+ DataflowPythonConfiguration,
+)
+from airflow.utils.decorators import apply_defaults
+from airflow.version import version
+
+
+class BeamRunPythonPipelineOperator(BaseOperator):
+ """
+ Launching Apache Beam pipelines written in Python. Note that both
+ ``default_pipeline_options`` and ``pipeline_options`` will be merged to specify pipeline
+ execution parameter, and ``default_pipeline_options`` is expected to save
+ high-level options, for instances, project and zone information, which
+ apply to all beam operators in the DAG.
+
+ .. code-block:: python
+
+ default_args = {
+ 'default_pipeline_options':
+ {
+ 'labels': 'example-label'
+ }
+ }
+
+ with models.DAG(
+ "example_beam_native_python",
+ default_args=default_args,
+ start_date=days_ago(1),
+ schedule_interval=None,
+ tags=['example'],
+ ) as dag_native_python:
+
+ start_python_job_local_direct_runner = BeamRunPythonPipelineOperator(
+ task_id="start_python_job_local_direct_runner",
+ runner="DirectRunner",
+ py_file='apache_beam.examples.wordcount',
+ py_options=['-m'],
+ py_requirements=['apache-beam[gcp]==2.21.0'],
+ py_interpreter='python3',
+ py_system_site_packages=False,
+ )
+
+ .. seealso::
+ For more detail on Apache Beam have a look at the reference:
+ https://beam.apache.org/documentation/
+
+ :param py_file: Reference to the python Apache Beam pipeline file.py, e.g.,
+ /some/local/file/path/to/your/python/pipeline/file. (templated)
+ :type py_file: str
+ :param runner: Runner on which pipeline will be run. By default "DirectRunner" is being used.
+ Other possible options: DataflowRunner, SparkRunner, FlinkRunner.
+ See: :class:`~providers.apache.beam.hooks.beam.BeamRunnerType`
+ See: https://beam.apache.org/documentation/runners/capability-matrix/
+
+ If you use Dataflow runner check dedicated operator:
+ :class:`~providers.google.cloud.operators.dataflow.DataflowCreatePythonJobOperator`
+ :type runner: str
+ :param py_options: Additional python options, e.g., ["-m", "-v"].
+ :type py_options: list[str]
+ :param default_pipeline_options: Map of default pipeline options.
+ :type default_pipeline_options: dict
+ :param pipeline_options: Map of pipeline 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.
+ :type pipeline_options: dict
+ :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
+ :type py_interpreter: str
+ :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.
+ :type py_requirements: List[str]
+ :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: Optional.
+ The connection ID to use connecting to Google Cloud Storage if pyfile is on GCS.
+ :type gcp_conn_id: str
+ :param delegate_to: Optional.
+ 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.
+ :type delegate_to: str
+ :param dataflow_config: Dataflow configuration, used when runner type is set to DataflowRunner
+ :type dataflow_config: Union[dict, providers.google.cloud.operators.dataflow.DataflowPythonConfiguration]
+ """
+
+ template_fields = ["py_file", "runner", "pipeline_options", "default_pipeline_options", "dataflow_config"]
+ template_fields_renderers = {'dataflow_config': 'json'}
+
+ @apply_defaults
+ def __init__(
+ self,
+ *,
+ py_file: str,
+ runner: str = "DirectRunner",
+ default_pipeline_options: Optional[dict] = None,
+ pipeline_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,
+ gcp_conn_id: str = "google_cloud_default",
+ delegate_to: Optional[str] = None,
+ dataflow_config: Optional[Union[DataflowPythonConfiguration, dict]] = None,
+ **kwargs,
+ ) -> None:
+ super().__init__(**kwargs)
+
+ self.py_file = py_file
+ self.runner = runner
+ self.py_options = py_options or []
+ self.default_pipeline_options = default_pipeline_options or {}
+ self.pipeline_options = pipeline_options or {}
+ self.pipeline_options.setdefault("labels", {}).update(
+ {"airflow-version": "v" + version.replace(".", "-").replace("+", "-")}
Review comment:
why not use version as is, instead of adding a "v" and changing . and +'s to dashes?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org