Source code for airflow.providers.google.cloud.operators.datafusion

# 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 Google DataFusion operators."""
from __future__ import annotations

import time
from typing import TYPE_CHECKING, Any, Sequence

from google.api_core.retry import exponential_sleep_generator
from googleapiclient.errors import HttpError

from airflow.configuration import conf
from airflow.exceptions import AirflowException, AirflowProviderDeprecationWarning
from airflow.providers.google.cloud.hooks.datafusion import SUCCESS_STATES, DataFusionHook, PipelineStates
from airflow.providers.google.cloud.links.datafusion import (
    DataFusionInstanceLink,
    DataFusionPipelineLink,
    DataFusionPipelinesLink,
)
from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
from airflow.providers.google.cloud.triggers.datafusion import DataFusionStartPipelineTrigger
from airflow.providers.google.cloud.utils.datafusion import DataFusionPipelineType
from airflow.providers.google.cloud.utils.helpers import resource_path_to_dict

if TYPE_CHECKING:
    from airflow.utils.context import Context


[docs]class DataFusionPipelineLinkHelper: """ Helper class for Pipeline links. .. warning:: This class is deprecated. Consider using ``resource_path_to_dict()`` instead. """ @staticmethod
[docs] def get_project_id(instance): raise AirflowProviderDeprecationWarning( "DataFusionPipelineLinkHelper is deprecated. Consider using resource_path_to_dict() instead." ) instance = instance["name"] project_id = next(x for x in instance.split("/") if x.startswith("airflow")) return project_id
[docs]class CloudDataFusionRestartInstanceOperator(GoogleCloudBaseOperator): """ Restart a single Data Fusion instance. At the end of an operation instance is fully restarted. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionRestartInstanceOperator` :param instance_name: The name of the instance to restart. :param location: The Cloud Data Fusion location in which to handle the request. :param project_id: The ID of the Google Cloud project that the instance belongs to. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "impersonation_chain", )
def __init__( self, *, instance_name: str, location: str, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Restarting Data Fusion instance: %s", self.instance_name) operation = hook.restart_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) instance = hook.wait_for_operation(operation) self.log.info("Instance %s restarted successfully", self.instance_name) project_id = resource_path_to_dict(resource_name=instance["name"])["projects"] DataFusionInstanceLink.persist( context=context, task_instance=self, project_id=project_id, instance_name=self.instance_name, location=self.location, )
[docs]class CloudDataFusionDeleteInstanceOperator(GoogleCloudBaseOperator): """ Deletes a single Date Fusion instance. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionDeleteInstanceOperator` :param instance_name: The name of the instance to restart. :param location: The Cloud Data Fusion location in which to handle the request. :param project_id: The ID of the Google Cloud project that the instance belongs to. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "impersonation_chain", )
def __init__( self, *, instance_name: str, location: str, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Deleting Data Fusion instance: %s", self.instance_name) operation = hook.delete_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) hook.wait_for_operation(operation) self.log.info("Instance %s deleted successfully", self.instance_name)
[docs]class CloudDataFusionCreateInstanceOperator(GoogleCloudBaseOperator): """ Creates a new Data Fusion instance in the specified project and location. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionCreateInstanceOperator` :param instance_name: The name of the instance to create. :param instance: An instance of Instance. https://cloud.google.com/data-fusion/docs/reference/rest/v1beta1/projects.locations.instances#Instance :param location: The Cloud Data Fusion location in which to handle the request. :param project_id: The ID of the Google Cloud project that the instance belongs to. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "instance", "impersonation_chain", )
def __init__( self, *, instance_name: str, instance: dict[str, Any], location: str, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.instance_name = instance_name self.instance = instance self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> dict: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Creating Data Fusion instance: %s", self.instance_name) try: operation = hook.create_instance( instance_name=self.instance_name, instance=self.instance, location=self.location, project_id=self.project_id, ) instance = hook.wait_for_operation(operation) self.log.info("Instance %s created successfully", self.instance_name) except HttpError as err: if err.resp.status not in (409, "409"): raise self.log.info("Instance %s already exists", self.instance_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id ) # Wait for instance to be ready for time_to_wait in exponential_sleep_generator(initial=10, maximum=120): if instance["state"] != "CREATING": break time.sleep(time_to_wait) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id ) project_id = resource_path_to_dict(resource_name=instance["name"])["projects"] DataFusionInstanceLink.persist( context=context, task_instance=self, project_id=project_id, instance_name=self.instance_name, location=self.location, ) return instance
[docs]class CloudDataFusionUpdateInstanceOperator(GoogleCloudBaseOperator): """ Updates a single Data Fusion instance. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionUpdateInstanceOperator` :param instance_name: The name of the instance to create. :param instance: An instance of Instance. https://cloud.google.com/data-fusion/docs/reference/rest/v1beta1/projects.locations.instances#Instance :param update_mask: Field mask is used to specify the fields that the update will overwrite in an instance resource. The fields specified in the updateMask are relative to the resource, not the full request. A field will be overwritten if it is in the mask. If the user does not provide a mask, all the supported fields (labels and options currently) will be overwritten. A comma-separated list of fully qualified names of fields. Example: "user.displayName,photo". https://developers.google.com/protocol-buffers/docs/reference/google.protobuf?_ga=2.205612571.-968688242.1573564810#google.protobuf.FieldMask :param location: The Cloud Data Fusion location in which to handle the request. :param project_id: The ID of the Google Cloud project that the instance belongs to. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "instance", "impersonation_chain", )
def __init__( self, *, instance_name: str, instance: dict[str, Any], update_mask: str, location: str, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.update_mask = update_mask self.instance_name = instance_name self.instance = instance self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Updating Data Fusion instance: %s", self.instance_name) operation = hook.patch_instance( instance_name=self.instance_name, instance=self.instance, update_mask=self.update_mask, location=self.location, project_id=self.project_id, ) instance = hook.wait_for_operation(operation) self.log.info("Instance %s updated successfully", self.instance_name) project_id = resource_path_to_dict(resource_name=instance["name"])["projects"] DataFusionInstanceLink.persist( context=context, task_instance=self, project_id=project_id, instance_name=self.instance_name, location=self.location, )
[docs]class CloudDataFusionGetInstanceOperator(GoogleCloudBaseOperator): """ Gets details of a single Data Fusion instance. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionGetInstanceOperator` :param instance_name: The name of the instance. :param location: The Cloud Data Fusion location in which to handle the request. :param project_id: The ID of the Google Cloud project that the instance belongs to. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "impersonation_chain", )
def __init__( self, *, instance_name: str, location: str, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> dict: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Retrieving Data Fusion instance: %s", self.instance_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) project_id = resource_path_to_dict(resource_name=instance["name"])["projects"] DataFusionInstanceLink.persist( context=context, task_instance=self, project_id=project_id, instance_name=self.instance_name, location=self.location, ) return instance
[docs]class CloudDataFusionCreatePipelineOperator(GoogleCloudBaseOperator): """ Creates a Cloud Data Fusion pipeline. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionCreatePipelineOperator` :param pipeline_name: Your pipeline name. :param pipeline: The pipeline definition. For more information check: https://docs.cdap.io/cdap/current/en/developer-manual/pipelines/developing-pipelines.html#pipeline-configuration-file-format :param instance_name: The name of the instance. :param location: The Cloud Data Fusion location in which to handle the request. :param namespace: If your pipeline belongs to a Basic edition instance, the namespace ID is always default. If your pipeline belongs to an Enterprise edition instance, you can create a namespace. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "pipeline_name", "impersonation_chain", )
def __init__( self, *, pipeline_name: str, pipeline: dict[str, Any], instance_name: str, location: str, namespace: str = "default", project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.pipeline_name = pipeline_name self.pipeline = pipeline self.namespace = namespace self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Creating Data Fusion pipeline: %s", self.pipeline_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) api_url = instance["apiEndpoint"] hook.create_pipeline( pipeline_name=self.pipeline_name, pipeline=self.pipeline, instance_url=api_url, namespace=self.namespace, ) DataFusionPipelineLink.persist( context=context, task_instance=self, uri=instance["serviceEndpoint"], pipeline_name=self.pipeline_name, ) self.log.info("Pipeline %s created", self.pipeline_name)
[docs]class CloudDataFusionDeletePipelineOperator(GoogleCloudBaseOperator): """ Deletes a Cloud Data Fusion pipeline. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionDeletePipelineOperator` :param pipeline_name: Your pipeline name. :param version_id: Version of pipeline to delete :param instance_name: The name of the instance. :param location: The Cloud Data Fusion location in which to handle the request. :param namespace: If your pipeline belongs to a Basic edition instance, the namespace ID is always default. If your pipeline belongs to an Enterprise edition instance, you can create a namespace. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "version_id", "pipeline_name", "impersonation_chain", )
def __init__( self, *, pipeline_name: str, instance_name: str, location: str, version_id: str | None = None, namespace: str = "default", project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.pipeline_name = pipeline_name self.version_id = version_id self.namespace = namespace self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Deleting Data Fusion pipeline: %s", self.pipeline_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) api_url = instance["apiEndpoint"] hook.delete_pipeline( pipeline_name=self.pipeline_name, version_id=self.version_id, instance_url=api_url, namespace=self.namespace, ) self.log.info("Pipeline deleted")
[docs]class CloudDataFusionListPipelinesOperator(GoogleCloudBaseOperator): """ Lists Cloud Data Fusion pipelines. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionListPipelinesOperator` :param instance_name: The name of the instance. :param location: The Cloud Data Fusion location in which to handle the request. :param artifact_version: Artifact version to filter instances :param artifact_name: Artifact name to filter instances :param namespace: If your pipeline belongs to a Basic edition instance, the namespace ID is always default. If your pipeline belongs to an Enterprise edition instance, you can create a namespace. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "artifact_name", "artifact_version", "impersonation_chain", )
def __init__( self, *, instance_name: str, location: str, artifact_name: str | None = None, artifact_version: str | None = None, namespace: str = "default", project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.artifact_version = artifact_version self.artifact_name = artifact_name self.namespace = namespace self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> dict: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Listing Data Fusion pipelines") instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) api_url = instance["apiEndpoint"] service_endpoint = instance["serviceEndpoint"] pipelines = hook.list_pipelines( instance_url=api_url, namespace=self.namespace, artifact_version=self.artifact_version, artifact_name=self.artifact_name, ) self.log.info("Pipelines: %s", pipelines) DataFusionPipelinesLink.persist(context=context, task_instance=self, uri=service_endpoint) return pipelines
[docs]class CloudDataFusionStartPipelineOperator(GoogleCloudBaseOperator): """ Starts a Cloud Data Fusion pipeline. Works for both batch and stream pipelines. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionStartPipelineOperator` :param pipeline_name: Your pipeline name. :param pipeline_type: Optional pipeline type (BATCH by default). :param instance_name: The name of the instance. :param success_states: If provided the operator will wait for pipeline to be in one of the provided states. :param pipeline_timeout: How long (in seconds) operator should wait for the pipeline to be in one of ``success_states``. Works only if ``success_states`` are provided. :param location: The Cloud Data Fusion location in which to handle the request. :param runtime_args: Optional runtime args to be passed to the pipeline :param namespace: If your pipeline belongs to a Basic edition instance, the namespace ID is always default. If your pipeline belongs to an Enterprise edition instance, you can create a namespace. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). :param asynchronous: Flag to return after submitting the pipeline ID to the Data Fusion API. This is useful for submitting long-running pipelines and waiting on them asynchronously using the CloudDataFusionPipelineStateSensor :param deferrable: Run operator in the deferrable mode. Is not related to asynchronous parameter. While asynchronous parameter gives a possibility to wait until pipeline reaches terminate state using sleep() method, deferrable mode checks for the state using asynchronous calls. It is not possible to use both asynchronous and deferrable parameters at the same time. :param poll_interval: Polling period in seconds to check for the status. Used only in deferrable mode. """
[docs] template_fields: Sequence[str] = ( "instance_name", "pipeline_name", "runtime_args", "impersonation_chain", )
def __init__( self, *, pipeline_name: str, instance_name: str, location: str, pipeline_type: DataFusionPipelineType = DataFusionPipelineType.BATCH, runtime_args: dict[str, Any] | None = None, success_states: list[str] | None = None, namespace: str = "default", pipeline_timeout: int = 5 * 60, project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, asynchronous=False, deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False), poll_interval=3.0, **kwargs, ) -> None: super().__init__(**kwargs) self.pipeline_name = pipeline_name self.pipeline_type = pipeline_type self.runtime_args = runtime_args self.namespace = namespace self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain self.asynchronous = asynchronous self.pipeline_timeout = pipeline_timeout self.deferrable = deferrable self.poll_interval = poll_interval if success_states: self.success_states = success_states else: self.success_states = [*SUCCESS_STATES, PipelineStates.RUNNING]
[docs] def execute(self, context: Context) -> str: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Starting Data Fusion pipeline: %s", self.pipeline_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) api_url = instance["apiEndpoint"] pipeline_id = hook.start_pipeline( pipeline_name=self.pipeline_name, pipeline_type=self.pipeline_type, instance_url=api_url, namespace=self.namespace, runtime_args=self.runtime_args, ) self.log.info("Pipeline %s submitted successfully.", pipeline_id) DataFusionPipelineLink.persist( context=context, task_instance=self, uri=instance["serviceEndpoint"], pipeline_name=self.pipeline_name, ) if self.deferrable: if self.asynchronous: raise AirflowException( "Both asynchronous and deferrable parameters were passed. Please, provide only one." ) self.defer( trigger=DataFusionStartPipelineTrigger( success_states=self.success_states, instance_url=api_url, namespace=self.namespace, pipeline_name=self.pipeline_name, pipeline_type=self.pipeline_type.value, pipeline_id=pipeline_id, poll_interval=self.poll_interval, gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain, ), method_name="execute_complete", ) else: if not self.asynchronous: # when NOT using asynchronous mode it will just wait for pipeline to finish and print message self.log.info("Waiting when pipeline %s will be in one of the success states", pipeline_id) hook.wait_for_pipeline_state( success_states=self.success_states, pipeline_id=pipeline_id, pipeline_name=self.pipeline_name, pipeline_type=self.pipeline_type, namespace=self.namespace, instance_url=api_url, timeout=self.pipeline_timeout, ) self.log.info("Pipeline %s discovered success state.", pipeline_id) # otherwise, return pipeline_id so that sensor can use it later to check the pipeline state return pipeline_id
[docs] def execute_complete(self, context: Context, event: dict[str, Any]): """ Callback for when the trigger fires - returns immediately. Relies on trigger to throw an exception, otherwise it assumes execution was successful. """ if event["status"] == "error": raise AirflowException(event["message"]) self.log.info( "%s completed with response %s ", self.task_id, event["message"], ) return event["pipeline_id"]
[docs]class CloudDataFusionStopPipelineOperator(GoogleCloudBaseOperator): """ Stops a Cloud Data Fusion pipeline. Works for both batch and stream pipelines. .. seealso:: For more information on how to use this operator, take a look at the guide: :ref:`howto/operator:CloudDataFusionStopPipelineOperator` :param pipeline_name: Your pipeline name. :param instance_name: The name of the instance. :param location: The Cloud Data Fusion location in which to handle the request. :param namespace: If your pipeline belongs to a Basic edition instance, the namespace ID is always default. If your pipeline belongs to an Enterprise edition instance, you can create a namespace. :param api_version: The version of the api that will be requested for example 'v3'. :param gcp_conn_id: The connection ID to use when fetching connection info. :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). """
[docs] template_fields: Sequence[str] = ( "instance_name", "pipeline_name", "impersonation_chain", )
def __init__( self, *, pipeline_name: str, instance_name: str, location: str, namespace: str = "default", project_id: str | None = None, api_version: str = "v1beta1", gcp_conn_id: str = "google_cloud_default", impersonation_chain: str | Sequence[str] | None = None, **kwargs, ) -> None: super().__init__(**kwargs) self.pipeline_name = pipeline_name self.namespace = namespace self.instance_name = instance_name self.location = location self.project_id = project_id self.api_version = api_version self.gcp_conn_id = gcp_conn_id self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context) -> None: hook = DataFusionHook( gcp_conn_id=self.gcp_conn_id, api_version=self.api_version, impersonation_chain=self.impersonation_chain, ) self.log.info("Data Fusion pipeline: %s is going to be stopped", self.pipeline_name) instance = hook.get_instance( instance_name=self.instance_name, location=self.location, project_id=self.project_id, ) api_url = instance["apiEndpoint"] DataFusionPipelineLink.persist( context=context, task_instance=self, uri=instance["serviceEndpoint"], pipeline_name=self.pipeline_name, ) hook.stop_pipeline( pipeline_name=self.pipeline_name, instance_url=api_url, namespace=self.namespace, ) self.log.info("Pipeline stopped")

Was this entry helpful?