To set up a Lifecycle configurations, open the Sagemaker panel and navigate to Lifecycle configurations under the section Admin configurations.

Select the Notebook Instance tab and click the Create configuration button

Paste the below code to the entry area.

#!/bin/bash

set -e
sudo -u ec2-user -i <<'EOF'
# 1. Retraining and redeploying the model
NOTEBOOK_FILE=/home/ec2-user/SageMaker/tyre_quality_classification_local_training.ipynb
echo "Activating conda env"
source /home/ec2-user/anaconda3/bin/activate pytorch_p310
nohup jupyter nbconvert "$NOTEBOOK_FILE" --ExecutePreprocessor.kernel_name=python --execute --to notebook &
nbconvert_pid=$!
conda deactivate

# 2. Scheduling a job to shutdown the notebook to save the cost
PYTHON_DIR='/home/ec2-user/anaconda3/envs/JupyterSystemEnv/bin/python3.10'
echo "Starting the autostop script in cron"
(crontab -l 2>/dev/null; echo "*/5 * * * * bash -c 'if ps -p $nbconvert_pid > /dev/null; then echo \"Notebook is still running.\" >> /var/log/jupyter.log; else echo \"Notebook execution completed.\" >> /var/log/jupyter.log; $PYTHON_DIR -c \"import boto3;boto3.client(\'sagemaker\').stop_notebook_instance(NotebookInstanceName=get_notebook_name())\" >> /var/log/jupyter.log; fi'") | crontab -
EOF

This script executes the Jupyter Notebook, which handles the retraining and redeployment of the model for inference. After the execution is complete, the notebook will automatically shut down within 5 minutes. To learn more about the problem statement and the code implementation, please refer to Part 2 - Leveraging AWS FSx for NetApp ONTAP (FSxN) as a Data Source for Model Training in SageMaker.

After the creation, navigate to Notebook instances, select the target instance, and click Update settings under Actions dropdown.

Select the created Lifecycle configuration and click Update notebook instance.

To create an AWS Lambda function, navigate to the respective panel, switch to the Functions tab, and click on Create Function.

Please file all required entries on the page and remember to switch the Runtime to Python 3.10.

Please verify that the designated role has the required permission AmazonSageMakerFullAccess and click on the Create function button.

Select the created Lambda function. In the code tab, copy and paste the following code into the text area. This code starts the notebook instance named fsxn-ontap.

import boto3
import logging

def lambda_handler(event, context):
    client = boto3.client('sagemaker')
    logging.info('Invoking SageMaker')
    client.start_notebook_instance(NotebookInstanceName='fsxn-ontap')
    return {
        'statusCode': 200,
        'body': f'Starting notebook instance: {notebook_instance_name}'
    }

Click the Deploy button to apply this code change.

To specify how to trigger this AWS Lambda function, click on the Add Trigger button.

Select EventBridge from the dropdown menu, then click on the radio button labeled Create a new rule. In the schedule expression field, enter rate(1 day), and click on the Add button to create and apply this new cron job rule to the AWS Lambda function.