Création d'une planification de surveillance avec une ressource personnalisée AWS CloudFormation - Amazon SageMaker

Création d'une planification de surveillance avec une ressource personnalisée AWS CloudFormation

Pour utiliser AWS CloudFormation afin de créer une planification de surveillance, utilisez une ressource personnalisée AWS CloudFormation. La ressource personnalisée se trouve dans Python. Pour la déployer, veuillez consulter Package de déploiement AWS Lambda dans Python.

Ressource personnalisée

Commencez par ajouter une ressource personnalisée à votre modèle AWS CloudFormation. Cela pointera vers une fonction AWS Lambda que vous créerez à l'étape suivante.

Cette ressource permet de personnaliser les paramètres du programme de surveillance. Vous pouvez ajouter ou supprimer d'autres paramètres en modifiant la ressource AWS CloudFormation et la fonction Lambda dans l'exemple de ressource suivant.

{ "AWSTemplateFormatVersion": "2010-09-09", "Resources": { "MonitoringSchedule": { "Type": "Custom::MonitoringSchedule", "Version": "1.0", "Properties": { "ServiceToken": "arn:aws:lambda:us-west-2:111111111111:function:lambda-name", "ScheduleName": "YourScheduleName", "EndpointName": "YourEndpointName", "BaselineConstraintsUri": "s3://your-baseline-constraints/constraints.json", "BaselineStatisticsUri": "s3://your-baseline-stats/statistics.json", "PostAnalyticsProcessorSourceUri": "s3://your-post-processor/postprocessor.py", "RecordPreprocessorSourceUri": "s3://your-preprocessor/preprocessor.py", "InputLocalPath": "/opt/ml/processing/endpointdata", "OutputLocalPath": "/opt/ml/processing/localpath", "OutputS3URI": "s3://your-output-uri", "ImageURI": "111111111111.dkr.ecr.us-west-2.amazonaws.com/your-image", "ScheduleExpression": "cron(0 * ? * * *)", "PassRoleArn": "arn:aws:iam::111111111111:role/AmazonSageMaker-ExecutionRole" } } } }

Code de ressource personnalisée Lambda

Cette ressource personnalisée AWS CloudFormation utilise la bibliothèque AWS Custom Resource Helper qui s'installe avec pip à l'aide de pip install crhelper.

Cette fonction Lambda est invoquée par AWS CloudFormation lors de la création et de la suppression de la pile. Cette fonction Lambda est responsable de la création et de la suppression de la planification de la surveillance et de l'utilisation des paramètres définis dans la ressource personnalisée décrite à la section précédente.

import boto3 import botocore import logging from crhelper import CfnResource from botocore.exceptions import ClientError logger = logging.getLogger(__name__) sm = boto3.client('sagemaker') # cfnhelper makes it easier to implement a CloudFormation custom resource helper = CfnResource() # CFN Handlers def handler(event, context): helper(event, context) @helper.create def create_handler(event, context): """ Called when CloudFormation custom resource sends the create event """ create_monitoring_schedule(event) @helper.delete def delete_handler(event, context): """ Called when CloudFormation custom resource sends the delete event """ schedule_name = get_schedule_name(event) delete_monitoring_schedule(schedule_name) @helper.poll_create def poll_create(event, context): """ Return true if the resource has been created and false otherwise so CloudFormation polls again. """ schedule_name = get_schedule_name(event) logger.info('Polling for creation of schedule: %s', schedule_name) return is_schedule_ready(schedule_name) @helper.update def noop(): """ Not currently implemented but crhelper will throw an error if it isn't added """ pass # Helper Functions def get_schedule_name(event): return event['ResourceProperties']['ScheduleName'] def create_monitoring_schedule(event): schedule_name = get_schedule_name(event) monitoring_schedule_config = create_monitoring_schedule_config(event) logger.info('Creating monitoring schedule with name: %s', schedule_name) sm.create_monitoring_schedule( MonitoringScheduleName=schedule_name, MonitoringScheduleConfig=monitoring_schedule_config) def is_schedule_ready(schedule_name): is_ready = False schedule = sm.describe_monitoring_schedule(MonitoringScheduleName=schedule_name) status = schedule['MonitoringScheduleStatus'] if status == 'Scheduled': logger.info('Monitoring schedule (%s) is ready', schedule_name) is_ready = True elif status == 'Pending': logger.info('Monitoring schedule (%s) still creating, waiting and polling again...', schedule_name) else: raise Exception('Monitoring schedule ({}) has unexpected status: {}'.format(schedule_name, status)) return is_ready def create_monitoring_schedule_config(event): props = event['ResourceProperties'] return { "ScheduleConfig": { "ScheduleExpression": props["ScheduleExpression"], }, "MonitoringJobDefinition": { "BaselineConfig": { "ConstraintsResource": { "S3Uri": props['BaselineConstraintsUri'], }, "StatisticsResource": { "S3Uri": props['BaselineStatisticsUri'], } }, "MonitoringInputs": [ { "EndpointInput": { "EndpointName": props["EndpointName"], "LocalPath": props["InputLocalPath"], } } ], "MonitoringOutputConfig": { "MonitoringOutputs": [ { "S3Output": { "S3Uri": props["OutputS3URI"], "LocalPath": props["OutputLocalPath"], } } ], }, "MonitoringResources": { "ClusterConfig": { "InstanceCount": 1, "InstanceType": "ml.t3.medium", "VolumeSizeInGB": 50, } }, "MonitoringAppSpecification": { "ImageUri": props["ImageURI"], "RecordPreprocessorSourceUri": props['PostAnalyticsProcessorSourceUri'], "PostAnalyticsProcessorSourceUri": props['PostAnalyticsProcessorSourceUri'], }, "StoppingCondition": { "MaxRuntimeInSeconds": 300 }, "RoleArn": props["PassRoleArn"], } } def delete_monitoring_schedule(schedule_name): logger.info('Deleting schedule: %s', schedule_name) try: sm.delete_monitoring_schedule(MonitoringScheduleName=schedule_name) except ClientError as e: if e.response['Error']['Code'] == 'ResourceNotFound': logger.info('Resource not found, nothing to delete') else: logger.error('Unexpected error while trying to delete monitoring schedule') raise e