wehub-resource-sync
682 lines
24 KiB
Python
682 lines
24 KiB
Python
import boto3
|
|
import pytest
|
|
|
|
from tests.helper_functions import set_boto_credentials # noqa: F401
|
|
from tests.sagemaker.mock import mock_sagemaker
|
|
|
|
|
|
@pytest.fixture
|
|
def sagemaker_client():
|
|
return boto3.client("sagemaker", region_name="us-west-2")
|
|
|
|
|
|
def create_sagemaker_model(sagemaker_client, model_name):
|
|
return sagemaker_client.create_model(
|
|
ExecutionRoleArn="arn:aws:iam::012345678910:role/sample-role",
|
|
ModelName=model_name,
|
|
PrimaryContainer={
|
|
"Image": "012345678910.dkr.ecr.us-west-2.amazonaws.com/sample-container",
|
|
},
|
|
)
|
|
|
|
|
|
def create_endpoint_config(sagemaker_client, endpoint_config_name, model_name):
|
|
return sagemaker_client.create_endpoint_config(
|
|
EndpointConfigName=endpoint_config_name,
|
|
ProductionVariants=[
|
|
{
|
|
"VariantName": "sample-variant",
|
|
"ModelName": model_name,
|
|
"InitialInstanceCount": 1,
|
|
"InstanceType": "ml.m4.xlarge",
|
|
"InitialVariantWeight": 1.0,
|
|
},
|
|
],
|
|
)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_created_model_is_listed_by_list_models_function(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
models_response = sagemaker_client.list_models()
|
|
assert "Models" in models_response
|
|
models = models_response["Models"]
|
|
assert all("ModelName" in model for model in models)
|
|
assert model_name in [model["ModelName"] for model in models]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_create_model_returns_arn_containing_model_name(sagemaker_client):
|
|
model_name = "sample-model"
|
|
model_create_response = create_sagemaker_model(
|
|
sagemaker_client=sagemaker_client, model_name=model_name
|
|
)
|
|
assert "ModelArn" in model_create_response
|
|
assert model_name in model_create_response["ModelArn"]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_creating_model_with_name_already_in_use_raises_exception(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
with pytest.raises(ValueError, match="Attempted to create a model"):
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_all_models_are_listed_after_creating_many_models(sagemaker_client):
|
|
model_names = []
|
|
|
|
for i in range(100):
|
|
model_name = f"sample-model-{i}"
|
|
model_names.append(model_name)
|
|
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
listed_models = sagemaker_client.list_models()["Models"]
|
|
listed_model_names = [model["ModelName"] for model in listed_models]
|
|
for model_name in model_names:
|
|
assert model_name in listed_model_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_model_response_contains_expected_attributes(sagemaker_client):
|
|
model_name = "sample-model"
|
|
execution_role_arn = "arn:aws:iam::012345678910:role/sample-role"
|
|
primary_container = {
|
|
"Image": "012345678910.dkr.ecr.us-west-2.amazonaws.com/sample-container",
|
|
}
|
|
|
|
sagemaker_client.create_model(
|
|
ModelName=model_name,
|
|
ExecutionRoleArn=execution_role_arn,
|
|
PrimaryContainer=primary_container,
|
|
)
|
|
|
|
describe_model_response = sagemaker_client.describe_model(ModelName=model_name)
|
|
assert "CreationTime" in describe_model_response
|
|
assert "ModelArn" in describe_model_response
|
|
assert "ExecutionRoleArn" in describe_model_response
|
|
assert describe_model_response["ExecutionRoleArn"] == execution_role_arn
|
|
assert "ModelName" in describe_model_response
|
|
assert describe_model_response["ModelName"] == model_name
|
|
assert "PrimaryContainer" in describe_model_response
|
|
assert describe_model_response["PrimaryContainer"] == primary_container
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_model_throws_exception_for_nonexistent_model(sagemaker_client):
|
|
with pytest.raises(ValueError, match="Attempted to describe a model"):
|
|
sagemaker_client.describe_model(ModelName="nonexistent-model")
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_model_is_no_longer_listed_after_deletion(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
sagemaker_client.delete_model(ModelName=model_name)
|
|
|
|
listed_models = sagemaker_client.list_models()["Models"]
|
|
listed_model_names = [model["ModelName"] for model in listed_models]
|
|
assert model_name not in listed_model_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_created_endpoint_config_is_listed_by_list_endpoints_function(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_configs_response = sagemaker_client.list_endpoint_configs()
|
|
assert "EndpointConfigs" in endpoint_configs_response
|
|
endpoint_configs = endpoint_configs_response["EndpointConfigs"]
|
|
assert all("EndpointConfigName" in endpoint_config for endpoint_config in endpoint_configs)
|
|
assert endpoint_config_name in [
|
|
endpoint_config["EndpointConfigName"] for endpoint_config in endpoint_configs
|
|
]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_create_endpoint_config_returns_arn_containing_config_name(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_config_response = create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
assert "EndpointConfigArn" in create_config_response
|
|
assert endpoint_config_name in create_config_response["EndpointConfigArn"]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_creating_endpoint_config_with_name_already_in_use_raises_exception(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
with pytest.raises(ValueError, match="Attempted to create an endpoint configuration"):
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_all_endpoint_configs_are_listed_after_creating_many_configs(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
endpoint_config_names = []
|
|
|
|
for i in range(100):
|
|
endpoint_config_name = f"sample-config-{i}"
|
|
endpoint_config_names.append(endpoint_config_name)
|
|
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
listed_endpoint_configs = sagemaker_client.list_endpoint_configs()["EndpointConfigs"]
|
|
listed_endpoint_config_names = [
|
|
endpoint_config["EndpointConfigName"] for endpoint_config in listed_endpoint_configs
|
|
]
|
|
for endpoint_config_name in endpoint_config_names:
|
|
assert endpoint_config_name in listed_endpoint_config_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_endpoint_config_response_contains_expected_attributes(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
production_variants = [
|
|
{
|
|
"VariantName": "sample-variant",
|
|
"ModelName": model_name,
|
|
"InitialInstanceCount": 1,
|
|
"InstanceType": "ml.m4.xlarge",
|
|
"InitialVariantWeight": 1.0,
|
|
},
|
|
]
|
|
async_inference_config = {
|
|
"ClientConfig": {"MaxConcurrentInvocationsPerInstance": 4},
|
|
"OutputConfig": {"S3OutputPath": "s3://bucket_name/", "NotificationConfig": {}},
|
|
}
|
|
sagemaker_client.create_endpoint_config(
|
|
EndpointConfigName=endpoint_config_name,
|
|
ProductionVariants=production_variants,
|
|
AsyncInferenceConfig=async_inference_config,
|
|
)
|
|
|
|
describe_endpoint_config_response = sagemaker_client.describe_endpoint_config(
|
|
EndpointConfigName=endpoint_config_name
|
|
)
|
|
assert "CreationTime" in describe_endpoint_config_response
|
|
assert "EndpointConfigArn" in describe_endpoint_config_response
|
|
assert "EndpointConfigName" in describe_endpoint_config_response
|
|
assert describe_endpoint_config_response["EndpointConfigName"] == endpoint_config_name
|
|
assert "ProductionVariants" in describe_endpoint_config_response
|
|
assert describe_endpoint_config_response["ProductionVariants"] == production_variants
|
|
assert "AsyncInferenceConfig" in describe_endpoint_config_response
|
|
assert describe_endpoint_config_response["AsyncInferenceConfig"] == async_inference_config
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_endpoint_config_throws_exception_for_nonexistent_config(sagemaker_client):
|
|
with pytest.raises(ValueError, match="Attempted to describe an endpoint config"):
|
|
sagemaker_client.describe_endpoint_config(EndpointConfigName="nonexistent-config")
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_endpoint_config_is_no_longer_listed_after_deletion(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
sagemaker_client.delete_endpoint_config(EndpointConfigName=endpoint_config_name)
|
|
|
|
listed_endpoint_configs = sagemaker_client.list_endpoint_configs()["EndpointConfigs"]
|
|
listed_endpoint_config_names = [
|
|
endpoint_config["EndpointConfigName"] for endpoint_config in listed_endpoint_configs
|
|
]
|
|
assert endpoint_config_name not in listed_endpoint_config_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_created_endpoint_is_listed_by_list_endpoints_function(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
endpoints_response = sagemaker_client.list_endpoints()
|
|
assert "Endpoints" in endpoints_response
|
|
endpoints = endpoints_response["Endpoints"]
|
|
assert all("EndpointName" in endpoint for endpoint in endpoints)
|
|
assert endpoint_name in [endpoint["EndpointName"] for endpoint in endpoints]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_create_endpoint_returns_arn_containing_endpoint_name(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
|
|
create_endpoint_response = sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
assert "EndpointArn" in create_endpoint_response
|
|
assert endpoint_name in create_endpoint_response["EndpointArn"]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_creating_endpoint_with_name_already_in_use_raises_exception(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
with pytest.raises(ValueError, match="Attempted to create an endpoint"):
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_all_endpoint_are_listed_after_creating_many_endpoints(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_names = []
|
|
|
|
for i in range(100):
|
|
endpoint_name = f"sample-endpoint-{i}"
|
|
endpoint_names.append(endpoint_name)
|
|
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
listed_endpoints = sagemaker_client.list_endpoints()["Endpoints"]
|
|
listed_endpoint_names = [endpoint["EndpointName"] for endpoint in listed_endpoints]
|
|
for endpoint_name in endpoint_names:
|
|
assert endpoint_name in listed_endpoint_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_endpoint_response_contains_expected_attributes(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
production_variants = [
|
|
{
|
|
"VariantName": "sample-variant",
|
|
"ModelName": model_name,
|
|
"InitialInstanceCount": 1,
|
|
"InstanceType": "ml.m4.xlarge",
|
|
"InitialVariantWeight": 1.0,
|
|
},
|
|
]
|
|
sagemaker_client.create_endpoint_config(
|
|
EndpointConfigName=endpoint_config_name,
|
|
ProductionVariants=production_variants,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
sagemaker_client.create_endpoint(
|
|
EndpointName=endpoint_name,
|
|
EndpointConfigName=endpoint_config_name,
|
|
)
|
|
|
|
describe_endpoint_response = sagemaker_client.describe_endpoint(EndpointName=endpoint_name)
|
|
assert "CreationTime" in describe_endpoint_response
|
|
assert "LastModifiedTime" in describe_endpoint_response
|
|
assert "EndpointArn" in describe_endpoint_response
|
|
assert "EndpointStatus" in describe_endpoint_response
|
|
assert "ProductionVariants" in describe_endpoint_response
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_endpoint_throws_exception_for_nonexistent_endpoint(sagemaker_client):
|
|
with pytest.raises(ValueError, match="Attempted to describe an endpoint"):
|
|
sagemaker_client.describe_endpoint(EndpointName="nonexistent-endpoint")
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_endpoint_is_no_longer_listed_after_deletion(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
)
|
|
|
|
sagemaker_client.delete_endpoint(EndpointName=endpoint_name)
|
|
|
|
listed_endpoints = sagemaker_client.list_endpoints()["Endpoints"]
|
|
listed_endpoint_names = [endpoint["EndpointName"] for endpoint in listed_endpoints]
|
|
assert endpoint_name not in listed_endpoint_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_update_endpoint_modifies_config_correctly(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
first_endpoint_config_name = "sample-config-1"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=first_endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
second_endpoint_config_name = "sample-config-2"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=second_endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=first_endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
)
|
|
|
|
first_describe_endpoint_response = sagemaker_client.describe_endpoint(
|
|
EndpointName=endpoint_name
|
|
)
|
|
assert first_describe_endpoint_response["EndpointConfigName"] == first_endpoint_config_name
|
|
|
|
sagemaker_client.update_endpoint(
|
|
EndpointName=endpoint_name, EndpointConfigName=second_endpoint_config_name
|
|
)
|
|
|
|
second_describe_endpoint_response = sagemaker_client.describe_endpoint(
|
|
EndpointName=endpoint_name
|
|
)
|
|
assert second_describe_endpoint_response["EndpointConfigName"] == second_endpoint_config_name
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_update_endpoint_with_nonexistent_config_throws_exception(sagemaker_client):
|
|
model_name = "sample-model-name"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
endpoint_config_name = "sample-config"
|
|
create_endpoint_config(
|
|
sagemaker_client=sagemaker_client,
|
|
endpoint_config_name=endpoint_config_name,
|
|
model_name=model_name,
|
|
)
|
|
|
|
endpoint_name = "sample-endpoint"
|
|
sagemaker_client.create_endpoint(
|
|
EndpointConfigName=endpoint_config_name,
|
|
EndpointName=endpoint_name,
|
|
)
|
|
|
|
with pytest.raises(ValueError, match="Attempted to update an endpoint"):
|
|
sagemaker_client.update_endpoint(
|
|
EndpointName=endpoint_name, EndpointConfigName="nonexistent-config"
|
|
)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_created_transform_job_is_listed_by_list_transform_jobs_function(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
transform_input = {
|
|
"DataSource": {"S3DataSource": {"S3DataType": "Some Data Type", "S3Uri": "Some Input Uri"}}
|
|
}
|
|
|
|
transform_output = {"S3OutputPath": "Some Output Path"}
|
|
|
|
transform_resources = {"InstanceType": "Some Instance Type", "InstanceCount": 1}
|
|
|
|
job_name = "sample-job"
|
|
|
|
sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
transform_jobs_response = sagemaker_client.list_transform_jobs()
|
|
assert "TransformJobSummaries" in transform_jobs_response
|
|
transform_jobs = transform_jobs_response["TransformJobSummaries"]
|
|
assert all("TransformJobName" in transform_job for transform_job in transform_jobs)
|
|
assert job_name in [transform_job["TransformJobName"] for transform_job in transform_jobs]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_create_transform_job_returns_arn_containing_transform_job_name(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
transform_input = {
|
|
"DataSource": {"S3DataSource": {"S3DataType": "Some Data Type", "S3Uri": "Some Input Uri"}}
|
|
}
|
|
|
|
transform_output = {"S3OutputPath": "Some Output Path"}
|
|
|
|
transform_resources = {"InstanceType": "Some Instance Type", "InstanceCount": 1}
|
|
|
|
job_name = "sample-job"
|
|
|
|
create_transform_job_response = sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
assert "TransformJobArn" in create_transform_job_response
|
|
assert job_name in create_transform_job_response["TransformJobArn"]
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_creating_transform_job_with_name_already_in_use_raises_exception(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
transform_input = {
|
|
"DataSource": {"S3DataSource": {"S3DataType": "Some Data Type", "S3Uri": "Some Input Uri"}}
|
|
}
|
|
|
|
transform_output = {"S3OutputPath": "Some Output Path"}
|
|
|
|
transform_resources = {"InstanceType": "Some Instance Type", "InstanceCount": 1}
|
|
|
|
job_name = "sample-job"
|
|
|
|
sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
with pytest.raises(ValueError, match="Attempted to create a transform job"):
|
|
sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_all_transform_jobs_are_listed_after_creating_many_transform_jobs(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
transform_input = {
|
|
"DataSource": {"S3DataSource": {"S3DataType": "Some Data Type", "S3Uri": "Some Input Uri"}}
|
|
}
|
|
|
|
transform_output = {"S3OutputPath": "Some Output Path"}
|
|
|
|
transform_resources = {"InstanceType": "Some Instance Type", "InstanceCount": 1}
|
|
|
|
job_names = []
|
|
|
|
for i in range(100):
|
|
job_name = f"sample-job-{i}"
|
|
job_names.append(job_name)
|
|
|
|
sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
listed_transform_jobs = sagemaker_client.list_transform_jobs()["TransformJobSummaries"]
|
|
listed_transform_job_names = [
|
|
transform_job["TransformJobName"] for transform_job in listed_transform_jobs
|
|
]
|
|
for job_name in job_names:
|
|
assert job_name in listed_transform_job_names
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_transform_job_response_contains_expected_attributes(sagemaker_client):
|
|
model_name = "sample-model"
|
|
create_sagemaker_model(sagemaker_client=sagemaker_client, model_name=model_name)
|
|
|
|
transform_input = {
|
|
"DataSource": {"S3DataSource": {"S3DataType": "Some Data Type", "S3Uri": "Some Input Uri"}}
|
|
}
|
|
|
|
transform_output = {"S3OutputPath": "Some Output Path"}
|
|
|
|
transform_resources = {"InstanceType": "Some Instance Type", "InstanceCount": 1}
|
|
|
|
job_name = "sample-job"
|
|
|
|
sagemaker_client.create_transform_job(
|
|
TransformJobName=job_name,
|
|
ModelName=model_name,
|
|
TransformInput=transform_input,
|
|
TransformOutput=transform_output,
|
|
TransformResources=transform_resources,
|
|
Tags=[{"Key": "Some Key", "Value": "Some Value"}],
|
|
)
|
|
|
|
describe_transform_job_response = sagemaker_client.describe_transform_job(
|
|
TransformJobName=job_name
|
|
)
|
|
assert "TransformJobName" in describe_transform_job_response
|
|
assert "CreationTime" in describe_transform_job_response
|
|
assert "TransformJobArn" in describe_transform_job_response
|
|
assert "TransformJobStatus" in describe_transform_job_response
|
|
assert "ModelName" in describe_transform_job_response
|
|
|
|
|
|
@mock_sagemaker
|
|
def test_describe_transform_job_throws_exception_for_nonexistent_transform_job(sagemaker_client):
|
|
with pytest.raises(ValueError, match="Attempted to describe a transform job"):
|
|
sagemaker_client.describe_transform_job(TransformJobName="nonexistent-job")
|