apache-airflow-providers-google 15.1.0rc1__py3-none-any.whl → 19.1.0rc1__py3-none-any.whl

airflow/providers/google/cloud/operators/cloud_memorystore.py

@@ -27,7 +27,7 @@ Operators for Google Cloud Memorystore service.
 from __future__ import annotations
 
 from collections.abc import Sequence
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any
 
 from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
 from google.cloud.memcache_v1beta2.types import cloud_memcache
@@ -50,7 +50,7 @@ if TYPE_CHECKING:
     from google.api_core.retry import Retry
     from google.protobuf.field_mask_pb2 import FieldMask
 
-    from airflow.utils.context import Context
+    from airflow.providers.common.compat.sdk import Context
 
 
 class CloudMemorystoreCreateInstanceOperator(GoogleCloudBaseOperator):
@@ -133,6 +133,13 @@ class CloudMemorystoreCreateInstanceOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance_id,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -148,9 +155,6 @@ class CloudMemorystoreCreateInstanceOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance_id,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
         return Instance.to_dict(result)
@@ -304,6 +308,13 @@ class CloudMemorystoreExportInstanceOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context) -> None:
         hook = CloudMemorystoreHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -320,9 +331,6 @@ class CloudMemorystoreExportInstanceOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
 
@@ -397,6 +405,13 @@ class CloudMemorystoreFailoverInstanceOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context) -> None:
         hook = CloudMemorystoreHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -412,9 +427,6 @@ class CloudMemorystoreFailoverInstanceOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
 
@@ -482,6 +494,13 @@ class CloudMemorystoreGetInstanceOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -496,9 +515,6 @@ class CloudMemorystoreGetInstanceOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
         return Instance.to_dict(result)
@@ -577,6 +593,13 @@ class CloudMemorystoreImportOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context) -> None:
         hook = CloudMemorystoreHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -592,9 +615,6 @@ class CloudMemorystoreImportOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
 
@@ -680,7 +700,6 @@ class CloudMemorystoreListInstancesOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceListLink.persist(
             context=context,
-            task_instance=self,
             project_id=self.project_id or hook.project_id,
         )
         instances = [Instance.to_dict(a) for a in result]
@@ -789,7 +808,6 @@ class CloudMemorystoreUpdateInstanceOperator(GoogleCloudBaseOperator):
         location_id, instance_id = res.name.split("/")[-3::2]
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
             instance_id=self.instance_id or instance_id,
             location_id=self.location or location_id,
             project_id=self.project_id or hook.project_id,
@@ -882,7 +900,6 @@ class CloudMemorystoreScaleInstanceOperator(GoogleCloudBaseOperator):
         location_id, instance_id = res.name.split("/")[-3::2]
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
             instance_id=self.instance_id or instance_id,
             location_id=self.location or location_id,
             project_id=self.project_id or hook.project_id,
@@ -1002,7 +1019,6 @@ class CloudMemorystoreCreateInstanceAndImportOperator(GoogleCloudBaseOperator):
         )
         RedisInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
             instance_id=self.instance_id,
             location_id=self.location,
             project_id=self.project_id or hook.project_id,
@@ -1171,6 +1187,14 @@ class CloudMemorystoreMemcachedApplyParametersOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance_id,
+            "location_id": self.location,
+            "project_id": self.project_id,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreMemcachedHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -1185,13 +1209,7 @@ class CloudMemorystoreMemcachedApplyParametersOperator(GoogleCloudBaseOperator):
             timeout=self.timeout,
             metadata=self.metadata,
         )
-        MemcachedInstanceDetailsLink.persist(
-            context=context,
-            task_instance=self,
-            instance_id=self.instance_id,
-            location_id=self.location,
-            project_id=self.project_id,
-        )
+        MemcachedInstanceDetailsLink.persist(context=context)
 
 
 class CloudMemorystoreMemcachedCreateInstanceOperator(GoogleCloudBaseOperator):
@@ -1263,6 +1281,13 @@ class CloudMemorystoreMemcachedCreateInstanceOperator(GoogleCloudBaseOperator):
         self.metadata = metadata
         self.gcp_conn_id = gcp_conn_id
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance_id,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreMemcachedHook(gcp_conn_id=self.gcp_conn_id)
         result = hook.create_instance(
@@ -1276,9 +1301,6 @@ class CloudMemorystoreMemcachedCreateInstanceOperator(GoogleCloudBaseOperator):
         )
         MemcachedInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance_id,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
         return cloud_memcache.Instance.to_dict(result)
@@ -1410,6 +1432,13 @@ class CloudMemorystoreMemcachedGetInstanceOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance,
+            "location_id": self.location,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreMemcachedHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -1424,9 +1453,6 @@ class CloudMemorystoreMemcachedGetInstanceOperator(GoogleCloudBaseOperator):
         )
         MemcachedInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
-            instance_id=self.instance,
-            location_id=self.location,
             project_id=self.project_id or hook.project_id,
         )
         return cloud_memcache.Instance.to_dict(result)
@@ -1506,7 +1532,6 @@ class CloudMemorystoreMemcachedListInstancesOperator(GoogleCloudBaseOperator):
         )
         MemcachedInstanceListLink.persist(
             context=context,
-            task_instance=self,
             project_id=self.project_id or hook.project_id,
         )
         instances = [cloud_memcache.Instance.to_dict(a) for a in result]
@@ -1612,7 +1637,6 @@ class CloudMemorystoreMemcachedUpdateInstanceOperator(GoogleCloudBaseOperator):
         location_id, instance_id = res.name.split("/")[-3::2]
         MemcachedInstanceDetailsLink.persist(
             context=context,
-            task_instance=self,
             instance_id=self.instance_id or instance_id,
             location_id=self.location or location_id,
             project_id=self.project_id or hook.project_id,
@@ -1688,6 +1712,14 @@ class CloudMemorystoreMemcachedUpdateParametersOperator(GoogleCloudBaseOperator)
         self.gcp_conn_id = gcp_conn_id
         self.impersonation_chain = impersonation_chain
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance_id": self.instance_id,
+            "location_id": self.location,
+            "project_id": self.project_id,
+        }
+
     def execute(self, context: Context):
         hook = CloudMemorystoreMemcachedHook(
             gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain
@@ -1702,10 +1734,4 @@ class CloudMemorystoreMemcachedUpdateParametersOperator(GoogleCloudBaseOperator)
             timeout=self.timeout,
             metadata=self.metadata,
         )
-        MemcachedInstanceDetailsLink.persist(
-            context=context,
-            task_instance=self,
-            instance_id=self.instance_id,
-            location_id=self.location,
-            project_id=self.project_id,
-        )
+        MemcachedInstanceDetailsLink.persist(context=context)

airflow/providers/google/cloud/operators/cloud_run.py

@@ -27,6 +27,7 @@ from google.cloud.run_v2 import Job, Service
 from airflow.configuration import conf
 from airflow.exceptions import AirflowException
 from airflow.providers.google.cloud.hooks.cloud_run import CloudRunHook, CloudRunServiceHook
+from airflow.providers.google.cloud.links.cloud_run import CloudRunJobLoggingLink
 from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
 from airflow.providers.google.cloud.triggers.cloud_run import CloudRunJobFinishedTrigger, RunJobStatus
 
@@ -34,7 +35,7 @@ if TYPE_CHECKING:
     from google.api_core import operation
     from google.cloud.run_v2.types import Execution
 
-    from airflow.utils.context import Context
+    from airflow.providers.common.compat.sdk import Context
 
 
 class CloudRunCreateJobOperator(GoogleCloudBaseOperator):
@@ -140,7 +141,7 @@ class CloudRunUpdateJobOperator(GoogleCloudBaseOperator):
 
 class CloudRunDeleteJobOperator(GoogleCloudBaseOperator):
     """
-    Deletes a job and wait for the the operation to be completed. Pushes the deleted job to xcom.
+    Deletes a job and wait for the operation to be completed. Pushes the deleted job to xcom.
 
     :param project_id: Required. The ID of the Google Cloud project that the service belongs to.
     :param region: Required. The ID of the Google Cloud region that the service belongs to.
@@ -248,7 +249,7 @@ class CloudRunExecuteJobOperator(GoogleCloudBaseOperator):
 
     :param project_id: Required. The ID of the Google Cloud project that the service belongs to.
     :param region: Required. The ID of the Google Cloud region that the service belongs to.
-    :param job_name: Required. The name of the job to update.
+    :param job_name: Required. The name of the job to execute.
     :param overrides: Optional map of override values.
     :param gcp_conn_id: The connection ID used to connect to Google Cloud.
     :param polling_period_seconds: Optional. Control the rate of the poll for the result of deferrable run.
@@ -265,7 +266,17 @@ class CloudRunExecuteJobOperator(GoogleCloudBaseOperator):
     :param deferrable: Run the operator in deferrable mode.
     """
 
-    template_fields = ("project_id", "region", "gcp_conn_id", "impersonation_chain", "job_name", "overrides")
+    operator_extra_links = (CloudRunJobLoggingLink(),)
+    template_fields = (
+        "project_id",
+        "region",
+        "gcp_conn_id",
+        "impersonation_chain",
+        "job_name",
+        "overrides",
+        "polling_period_seconds",
+        "timeout_seconds",
+    )
 
     def __init__(
         self,
@@ -303,6 +314,12 @@ class CloudRunExecuteJobOperator(GoogleCloudBaseOperator):
         if self.operation is None:
             raise AirflowException("Operation is None")
 
+        if self.operation.metadata.log_uri:
+            CloudRunJobLoggingLink.persist(
+                context=context,
+                log_uri=self.operation.metadata.log_uri,
+            )
+
         if not self.deferrable:
             result: Execution = self._wait_for_operation(self.operation)
             self._fail_if_execution_failed(result)
@@ -424,9 +441,10 @@ class CloudRunCreateServiceOperator(GoogleCloudBaseOperator):
                 self.service_name,
                 self.region,
             )
-            return hook.get_service(
+            service = hook.get_service(
                 service_name=self.service_name, region=self.region, project_id=self.project_id
             )
+            return Service.to_dict(service)
         except google.cloud.exceptions.GoogleCloudError as e:
             self.log.error("An error occurred. Exiting.")
             raise e

airflow/providers/google/cloud/operators/cloud_sql.py

@@ -28,7 +28,7 @@ from googleapiclient.errors import HttpError
 
 from airflow.configuration import conf
 from airflow.exceptions import AirflowException
-from airflow.hooks.base import BaseHook
+from airflow.providers.common.compat.sdk import BaseHook
 from airflow.providers.google.cloud.hooks.cloud_sql import CloudSQLDatabaseHook, CloudSQLHook
 from airflow.providers.google.cloud.links.cloud_sql import CloudSQLInstanceDatabaseLink, CloudSQLInstanceLink
 from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
@@ -39,8 +39,8 @@ from airflow.providers.google.common.links.storage import FileDetailsLink
 
 if TYPE_CHECKING:
     from airflow.models import Connection
+    from airflow.providers.common.compat.sdk import Context
     from airflow.providers.openlineage.extractors import OperatorLineage
-    from airflow.utils.context import Context
 
 
 SETTINGS = "settings"
@@ -285,6 +285,12 @@ class CloudSQLBaseOperator(GoogleCloudBaseOperator):
                 return False
             raise e
 
+    @property
+    def extra_links_params(self) -> dict[str, Any]:
+        return {
+            "instance": self.instance,
+        }
+
     def execute(self, context: Context):
         pass
 
@@ -384,8 +390,6 @@ class CloudSQLCreateInstanceOperator(CloudSQLBaseOperator):
 
         CloudSQLInstanceLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
 
@@ -479,8 +483,6 @@ class CloudSQLInstancePatchOperator(CloudSQLBaseOperator):
             )
         CloudSQLInstanceLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
 
@@ -714,8 +716,6 @@ class CloudSQLCreateInstanceDatabaseOperator(CloudSQLBaseOperator):
         )
         CloudSQLInstanceDatabaseLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
         if self._check_if_db_exists(database, hook):
@@ -822,8 +822,6 @@ class CloudSQLPatchInstanceDatabaseOperator(CloudSQLBaseOperator):
             )
         CloudSQLInstanceDatabaseLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
         return hook.patch_database(
@@ -1004,13 +1002,10 @@ class CloudSQLExportInstanceOperator(CloudSQLBaseOperator):
         )
         CloudSQLInstanceLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
         FileDetailsLink.persist(
             context=context,
-            task_instance=self,
             uri=self.body["exportContext"]["uri"][5:],
             project_id=self.project_id or hook.project_id,
         )
@@ -1147,13 +1142,10 @@ class CloudSQLImportInstanceOperator(CloudSQLBaseOperator):
         )
         CloudSQLInstanceLink.persist(
             context=context,
-            task_instance=self,
-            cloud_sql_instance=self.instance,
             project_id=self.project_id or hook.project_id,
         )
         FileDetailsLink.persist(
             context=context,
-            task_instance=self,
             uri=self.body["importContext"]["uri"][5:],
             project_id=self.project_id or hook.project_id,
         )

airflow/providers/google/cloud/operators/cloud_storage_transfer_service.py

@@ -65,13 +65,15 @@ from airflow.providers.google.cloud.links.cloud_storage_transfer import (
 )
 from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
 from airflow.providers.google.cloud.triggers.cloud_storage_transfer_service import (
+    CloudDataTransferServiceRunJobTrigger,
     CloudStorageTransferServiceCheckJobStatusTrigger,
 )
 from airflow.providers.google.cloud.utils.helpers import normalize_directory_path
 from airflow.providers.google.common.hooks.base_google import PROVIDE_PROJECT_ID
 
 if TYPE_CHECKING:
-    from airflow.utils.context import Context
+    from airflow.providers.common.compat.sdk import Context
+    from airflow.providers.openlineage.extractors import OperatorLineage
 
 
 class TransferJobPreprocessor:
@@ -96,8 +98,8 @@ class TransferJobPreprocessor:
 
         aws_hook = AwsBaseHook(self.aws_conn_id, resource_type="s3")
         aws_credentials = aws_hook.get_credentials()
-        aws_access_key_id = aws_credentials.access_key  # type: ignore[attr-defined]
-        aws_secret_access_key = aws_credentials.secret_key  # type: ignore[attr-defined]
+        aws_access_key_id = aws_credentials.access_key
+        aws_secret_access_key = aws_credentials.secret_key
         self.body[TRANSFER_SPEC][AWS_S3_DATA_SOURCE][AWS_ACCESS_KEY] = {
             ACCESS_KEY_ID: aws_access_key_id,
             SECRET_ACCESS_KEY: aws_secret_access_key,
@@ -468,6 +470,8 @@ class CloudDataTransferServiceRunJobOperator(GoogleCloudBaseOperator):
         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 timeout: Time to wait for the operation to end in seconds. Defaults to 60 seconds if not specified.
+    :param deferrable: Run operator in the deferrable mode.
     """
 
     # [START gcp_transfer_job_run_template_fields]
@@ -489,6 +493,8 @@ class CloudDataTransferServiceRunJobOperator(GoogleCloudBaseOperator):
         api_version: str = "v1",
         project_id: str = PROVIDE_PROJECT_ID,
         google_impersonation_chain: str | Sequence[str] | None = None,
+        timeout: float | None = None,
+        deferrable: bool = conf.getboolean("operators", "default_deferrable", fallback=False),
         **kwargs,
     ) -> None:
         super().__init__(**kwargs)
@@ -497,6 +503,8 @@ class CloudDataTransferServiceRunJobOperator(GoogleCloudBaseOperator):
         self.gcp_conn_id = gcp_conn_id
         self.api_version = api_version
         self.google_impersonation_chain = google_impersonation_chain
+        self.timeout = timeout
+        self.deferrable = deferrable
 
     def _validate_inputs(self) -> None:
         if not self.job_name:
@@ -518,8 +526,32 @@ class CloudDataTransferServiceRunJobOperator(GoogleCloudBaseOperator):
                 job_name=self.job_name,
             )
 
+        if self.deferrable:
+            self.defer(
+                timeout=timedelta(seconds=self.timeout or 60),
+                trigger=CloudDataTransferServiceRunJobTrigger(
+                    job_name=self.job_name,
+                    project_id=project_id,
+                    gcp_conn_id=self.gcp_conn_id,
+                    impersonation_chain=self.google_impersonation_chain,
+                ),
+                method_name="execute_complete",
+            )
+
         return hook.run_transfer_job(job_name=self.job_name, project_id=project_id)
 
+    def execute_complete(self, context: Context, event: dict[str, Any]) -> Any:
+        """
+        Act as a callback for when the trigger fires.
+
+        This returns immediately. It relies on trigger to throw an exception,
+        otherwise it assumes execution was successful.
+        """
+        if event["status"] == "error":
+            raise AirflowException(event["message"])
+
+        return event["job_result"]
+
 
 class CloudDataTransferServiceGetOperationOperator(GoogleCloudBaseOperator):
     """
@@ -586,7 +618,6 @@ class CloudDataTransferServiceGetOperationOperator(GoogleCloudBaseOperator):
         if project_id:
             CloudStorageTransferDetailsLink.persist(
                 context=context,
-                task_instance=self,
                 project_id=project_id,
                 operation_name=self.operation_name,
             )
@@ -663,7 +694,6 @@ class CloudDataTransferServiceListOperationsOperator(GoogleCloudBaseOperator):
         if project_id:
             CloudStorageTransferListLink.persist(
                 context=context,
-                task_instance=self,
                 project_id=project_id,
             )
 
@@ -966,6 +996,7 @@ class CloudDataTransferServiceS3ToGCSOperator(GoogleCloudBaseOperator):
         self.aws_role_arn = aws_role_arn
         self.deferrable = deferrable
         self._validate_inputs()
+        self._transfer_job: dict[str, Any] | None = None
 
     def _validate_inputs(self) -> None:
         if self.delete_job_after_completion and not self.wait:
@@ -980,19 +1011,18 @@ class CloudDataTransferServiceS3ToGCSOperator(GoogleCloudBaseOperator):
 
         TransferJobPreprocessor(body=body, aws_conn_id=self.aws_conn_id, default_schedule=True).process_body()
 
-        job = hook.create_transfer_job(body=body)
-
+        self._transfer_job = hook.create_transfer_job(body=body)
         if self.wait:
             if not self.deferrable:
-                hook.wait_for_transfer_job(job, timeout=self.timeout)
+                hook.wait_for_transfer_job(self._transfer_job, timeout=self.timeout)
                 if self.delete_job_after_completion:
-                    hook.delete_transfer_job(job_name=job[NAME], project_id=self.project_id)
+                    hook.delete_transfer_job(job_name=self._transfer_job[NAME], project_id=self.project_id)
             else:
                 self.defer(
                     timeout=timedelta(seconds=self.timeout or 60),
                     trigger=CloudStorageTransferServiceCheckJobStatusTrigger(
-                        job_name=job[NAME],
-                        project_id=job[PROJECT_ID],
+                        job_name=self._transfer_job[NAME],
+                        project_id=self._transfer_job[PROJECT_ID],
                         gcp_conn_id=self.gcp_conn_id,
                         impersonation_chain=self.google_impersonation_chain,
                     ),
@@ -1042,6 +1072,57 @@ class CloudDataTransferServiceS3ToGCSOperator(GoogleCloudBaseOperator):
 
         return body
 
+    def get_openlineage_facets_on_complete(self, task_instance) -> OperatorLineage | None:
+        """Provide OpenLineage OperatorLineage for the S3->GCS transfer."""
+        from airflow.providers.common.compat.openlineage.facet import Dataset
+        from airflow.providers.google.cloud.openlineage.facets import (
+            CloudStorageTransferJobFacet,
+            CloudStorageTransferRunFacet,
+        )
+        from airflow.providers.openlineage.extractors import OperatorLineage
+
+        input_ds = Dataset(
+            namespace=f"s3://{self.s3_bucket}",
+            name=normalize_directory_path(self.s3_path) or "",
+        )
+
+        output_ds = Dataset(
+            namespace=f"gs://{self.gcs_bucket}",
+            name=normalize_directory_path(self.gcs_path) or "",
+        )
+
+        job = self._transfer_job or {}
+        job_facet = CloudStorageTransferJobFacet(
+            jobName=job.get(NAME),
+            projectId=job.get(PROJECT_ID, self.project_id),
+            description=job.get(DESCRIPTION, self.description),
+            status=job.get(STATUS),
+            sourceBucket=job.get(TRANSFER_SPEC, {})
+            .get(AWS_S3_DATA_SOURCE, {})
+            .get(BUCKET_NAME, self.s3_bucket),
+            sourcePath=job.get(TRANSFER_SPEC, {}).get(AWS_S3_DATA_SOURCE, {}).get(PATH, self.s3_path),
+            targetBucket=job.get(TRANSFER_SPEC, {}).get(GCS_DATA_SINK, {}).get(BUCKET_NAME, self.gcs_bucket),
+            targetPath=job.get(TRANSFER_SPEC, {}).get(GCS_DATA_SINK, {}).get(PATH, self.gcs_path),
+            objectConditions=job.get(TRANSFER_SPEC, {}).get("objectConditions", self.object_conditions),
+            transferOptions=job.get(TRANSFER_SPEC, {}).get("transferOptions", self.transfer_options),
+            schedule=job.get(SCHEDULE, self.schedule),
+        )
+
+        run_facet = CloudStorageTransferRunFacet(
+            jobName=job.get(NAME),
+            wait=self.wait,
+            timeout=self.timeout,
+            deferrable=self.deferrable,
+            deleteJobAfterCompletion=self.delete_job_after_completion,
+        )
+
+        return OperatorLineage(
+            inputs=[input_ds],
+            outputs=[output_ds],
+            job_facets={"cloudStorageTransferJob": job_facet},
+            run_facets={"cloudStorageTransferRun": run_facet},
+        )
+
 
 class CloudDataTransferServiceGCSToGCSOperator(GoogleCloudBaseOperator):
     """