#
# 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 Google Vertex AI service testing Model Service operations.
"""
from __future__ import annotations
import os
from datetime import datetime
from google.cloud.aiplatform import schema
from google.protobuf.json_format import ParseDict
from google.protobuf.struct_pb2 import Value
from airflow.models.dag import DAG
from airflow.providers.google.cloud.operators.gcs import (
GCSCreateBucketOperator,
GCSDeleteBucketOperator,
GCSSynchronizeBucketsOperator,
)
from airflow.providers.google.cloud.operators.vertex_ai.custom_job import (
CreateCustomTrainingJobOperator,
DeleteCustomTrainingJobOperator,
)
from airflow.providers.google.cloud.operators.vertex_ai.dataset import (
CreateDatasetOperator,
DeleteDatasetOperator,
)
from airflow.providers.google.cloud.operators.vertex_ai.model_service import (
AddVersionAliasesOnModelOperator,
DeleteModelOperator,
DeleteModelVersionOperator,
DeleteVersionAliasesOnModelOperator,
ExportModelOperator,
GetModelOperator,
ListModelsOperator,
ListModelVersionsOperator,
SetDefaultVersionOnModelOperator,
UploadModelOperator,
)
from airflow.providers.google.cloud.transfers.gcs_to_local import GCSToLocalFilesystemOperator
from airflow.utils.trigger_rule import TriggerRule
[docs]
ENV_ID = os.environ.get("SYSTEM_TESTS_ENV_ID", "default")
[docs]
PROJECT_ID = os.environ.get("SYSTEM_TESTS_GCP_PROJECT", "default")
[docs]
DAG_ID = "vertex_ai_model_service_operations"
[docs]
TRAIN_DISPLAY_NAME = f"train-housing-custom-{ENV_ID}"
[docs]
MODEL_DISPLAY_NAME = f"custom-housing-model-{ENV_ID}"
[docs]
RESOURCE_DATA_BUCKET = "airflow-system-tests-resources"
[docs]
DATA_SAMPLE_GCS_BUCKET_NAME = f"bucket_{DAG_ID}_{ENV_ID}".replace("_", "-")
[docs]
STAGING_BUCKET = f"gs://{DATA_SAMPLE_GCS_BUCKET_NAME}"
[docs]
DATA_SAMPLE_GCS_OBJECT_NAME = "vertex-ai/california_housing_train.csv"
[docs]
TABULAR_DATASET = {
"display_name": f"tabular-dataset-{ENV_ID}",
"metadata_schema_uri": schema.dataset.metadata.tabular,
"metadata": ParseDict(
{
"input_config": {
"gcs_source": {"uri": [f"gs://{DATA_SAMPLE_GCS_BUCKET_NAME}/{DATA_SAMPLE_GCS_OBJECT_NAME}"]}
}
},
Value(),
),
}
[docs]
CONTAINER_URI = "us-docker.pkg.dev/vertex-ai/training/tf-cpu.2-2:latest"
# LOCAL_TRAINING_SCRIPT_PATH should be set for Airflow which is running on distributed system.
# For example in Composer the correct path is `gcs/data/california_housing_training_script.py`.
# Because `gcs/data/` is shared folder for Airflow's workers.
[docs]
IS_COMPOSER = bool(os.environ.get("COMPOSER_ENVIRONMENT", ""))
[docs]
LOCAL_TRAINING_SCRIPT_PATH = "gcs/data/california_housing_training_script.py" if IS_COMPOSER else ""
[docs]
MODEL_OUTPUT_CONFIG = {
"artifact_destination": {
"output_uri_prefix": STAGING_BUCKET,
},
"export_format_id": "custom-trained",
}
[docs]
MODEL_SERVING_CONTAINER_URI = "us-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-2:latest"
[docs]
MODEL_OBJ = {
"display_name": f"model-{ENV_ID}",
"artifact_uri": "{{ti.xcom_pull('custom_task')['artifactUri']}}",
"container_spec": {
"image_uri": MODEL_SERVING_CONTAINER_URI,
"command": [],
"args": [],
"env": [],
"ports": [],
"predict_route": "",
"health_route": "",
},
}
[docs]
MODEL_OBJ_V2 = {
"display_name": f"model-{ENV_ID}-v2",
"artifact_uri": "{{ti.xcom_pull('custom_task')['artifactUri']}}",
"container_spec": {
"image_uri": MODEL_SERVING_CONTAINER_URI,
"command": [],
"args": [],
"env": [],
"ports": [],
"predict_route": "",
"health_route": "",
},
}
with DAG(
DAG_ID,
schedule="@once",
start_date=datetime(2021, 1, 1),
catchup=False,
render_template_as_native_obj=True,
tags=["example", "vertex_ai", "model_service"],
) as dag:
[docs]
create_bucket = GCSCreateBucketOperator(
task_id="create_bucket",
bucket_name=DATA_SAMPLE_GCS_BUCKET_NAME,
storage_class="REGIONAL",
location=REGION,
)
move_data_files = GCSSynchronizeBucketsOperator(
task_id="move_files_to_bucket",
source_bucket=RESOURCE_DATA_BUCKET,
source_object="vertex-ai/california-housing-data",
destination_bucket=DATA_SAMPLE_GCS_BUCKET_NAME,
destination_object="vertex-ai",
recursive=True,
)
download_training_script_file = GCSToLocalFilesystemOperator(
task_id="download_training_script_file",
object_name="vertex-ai/california_housing_training_script.py",
bucket=DATA_SAMPLE_GCS_BUCKET_NAME,
filename=LOCAL_TRAINING_SCRIPT_PATH,
)
create_tabular_dataset = CreateDatasetOperator(
task_id="tabular_dataset",
dataset=TABULAR_DATASET,
region=REGION,
project_id=PROJECT_ID,
)
tabular_dataset_id = create_tabular_dataset.output["dataset_id"]
create_custom_training_job = CreateCustomTrainingJobOperator(
task_id="custom_task",
staging_bucket=f"gs://{DATA_SAMPLE_GCS_BUCKET_NAME}",
display_name=TRAIN_DISPLAY_NAME,
script_path=LOCAL_TRAINING_SCRIPT_PATH,
container_uri=CONTAINER_URI,
requirements=["gcsfs==0.7.1"],
model_serving_container_image_uri=MODEL_SERVING_CONTAINER_URI,
# run params
dataset_id=tabular_dataset_id,
replica_count=1,
model_display_name=MODEL_DISPLAY_NAME,
region=REGION,
project_id=PROJECT_ID,
)
model_id_v1 = create_custom_training_job.output["model_id"]
create_custom_training_job_v2 = CreateCustomTrainingJobOperator(
task_id="custom_task_v2",
staging_bucket=f"gs://{DATA_SAMPLE_GCS_BUCKET_NAME}",
display_name=TRAIN_DISPLAY_NAME,
script_path=LOCAL_TRAINING_SCRIPT_PATH,
container_uri=CONTAINER_URI,
requirements=["gcsfs==0.7.1"],
model_serving_container_image_uri=MODEL_SERVING_CONTAINER_URI,
parent_model=model_id_v1,
# run params
dataset_id=tabular_dataset_id,
replica_count=1,
model_display_name=MODEL_DISPLAY_NAME,
region=REGION,
project_id=PROJECT_ID,
)
model_id_v2 = create_custom_training_job_v2.output["model_id"]
# [START how_to_cloud_vertex_ai_get_model_operator]
get_model = GetModelOperator(
task_id="get_model", region=REGION, project_id=PROJECT_ID, model_id=model_id_v1
)
# [END how_to_cloud_vertex_ai_get_model_operator]
# [START how_to_cloud_vertex_ai_list_model_versions_operator]
list_model_versions = ListModelVersionsOperator(
task_id="list_model_versions", region=REGION, project_id=PROJECT_ID, model_id=model_id_v1
)
# [END how_to_cloud_vertex_ai_list_model_versions_operator]
# [START how_to_cloud_vertex_ai_set_version_as_default_operator]
set_default_version = SetDefaultVersionOnModelOperator(
task_id="set_default_version",
project_id=PROJECT_ID,
region=REGION,
model_id=model_id_v2,
)
# [END how_to_cloud_vertex_ai_set_version_as_default_operator]
# [START how_to_cloud_vertex_ai_add_version_aliases_operator]
add_version_alias = AddVersionAliasesOnModelOperator(
task_id="add_version_alias",
project_id=PROJECT_ID,
region=REGION,
version_aliases=["new-version", "beta"],
model_id=model_id_v2,
)
# [END how_to_cloud_vertex_ai_add_version_aliases_operator]
# [START how_to_cloud_vertex_ai_upload_model_operator]
upload_model = UploadModelOperator(
task_id="upload_model",
region=REGION,
project_id=PROJECT_ID,
model=MODEL_OBJ,
)
upload_model_v1 = upload_model.output["model_id"]
# [END how_to_cloud_vertex_ai_upload_model_operator]
upload_model_with_parent_model = UploadModelOperator(
task_id="upload_model_with_parent_model",
region=REGION,
project_id=PROJECT_ID,
model=MODEL_OBJ_V2,
parent_model=upload_model_v1,
)
# [START how_to_cloud_vertex_ai_export_model_operator]
export_model = ExportModelOperator(
task_id="export_model",
project_id=PROJECT_ID,
region=REGION,
model_id=upload_model.output["model_id"],
output_config=MODEL_OUTPUT_CONFIG,
)
# [END how_to_cloud_vertex_ai_export_model_operator]
# [START how_to_cloud_vertex_ai_delete_model_operator]
delete_model = DeleteModelOperator(
task_id="delete_model",
project_id=PROJECT_ID,
region=REGION,
model_id=upload_model.output["model_id"],
trigger_rule=TriggerRule.ALL_DONE,
)
# [END how_to_cloud_vertex_ai_delete_model_operator]
delete_model_with_parent_model = DeleteModelOperator(
task_id="delete_model_with_parent_model",
project_id=PROJECT_ID,
region=REGION,
model_id=upload_model_with_parent_model.output["model_id"],
trigger_rule=TriggerRule.ALL_DONE,
)
# [START how_to_cloud_vertex_ai_list_models_operator]
list_models = ListModelsOperator(
task_id="list_models",
region=REGION,
project_id=PROJECT_ID,
)
# [END how_to_cloud_vertex_ai_list_models_operator]
# [START how_to_cloud_vertex_ai_delete_version_aliases_operator]
delete_version_alias = DeleteVersionAliasesOnModelOperator(
task_id="delete_version_alias",
project_id=PROJECT_ID,
region=REGION,
version_aliases=["new-version"],
model_id=model_id_v2,
)
# [END how_to_cloud_vertex_ai_delete_version_aliases_operator]
# [START how_to_cloud_vertex_ai_delete_version_operator]
delete_model_version = DeleteModelVersionOperator(
task_id="delete_model_version",
project_id=PROJECT_ID,
region=REGION,
model_id=model_id_v1,
trigger_rule=TriggerRule.ALL_DONE,
)
# [END how_to_cloud_vertex_ai_delete_version_operator]
delete_custom_training_job = DeleteCustomTrainingJobOperator(
task_id="delete_custom_training_job",
training_pipeline_id="{{ task_instance.xcom_pull(task_ids='custom_task', key='training_id') }}",
custom_job_id="{{ task_instance.xcom_pull(task_ids='custom_task', key='custom_job_id') }}",
region=REGION,
project_id=PROJECT_ID,
trigger_rule=TriggerRule.ALL_DONE,
)
delete_tabular_dataset = DeleteDatasetOperator(
task_id="delete_tabular_dataset",
dataset_id=tabular_dataset_id,
region=REGION,
project_id=PROJECT_ID,
trigger_rule=TriggerRule.ALL_DONE,
)
delete_bucket = GCSDeleteBucketOperator(
task_id="delete_bucket",
bucket_name=DATA_SAMPLE_GCS_BUCKET_NAME,
trigger_rule=TriggerRule.ALL_DONE,
)
(
# TEST SETUP
create_bucket
>> move_data_files
>> download_training_script_file
>> create_tabular_dataset
>> create_custom_training_job
>> create_custom_training_job_v2
# TEST BODY
>> get_model
>> list_model_versions
>> set_default_version
>> add_version_alias
>> upload_model
>> upload_model_with_parent_model
>> export_model
>> delete_model
>> delete_model_with_parent_model
>> list_models
# TEST TEARDOWN
>> delete_version_alias
>> delete_model_version
>> delete_custom_training_job
>> delete_tabular_dataset
>> delete_bucket
)
# ### Everything below this line is not part of example ###
# ### Just for system tests purpose ###
from tests_common.test_utils.watcher import watcher
# This test needs watcher in order to properly mark success/failure
# when "tearDown" task with trigger rule is part of the DAG
list(dag.tasks) >> watcher()
from tests_common.test_utils.system_tests import get_test_run # noqa: E402
# Needed to run the example DAG with pytest (see: tests/system/README.md#run_via_pytest)
[docs]
test_run = get_test_run(dag)