In the following steps you will use AWS CloudFormation and Amazon Service Catalog to create a self-service mechanism to create secure data science environments. You will first deploy a CloudFormation template which provisions a shared service environment which hosts a PyPI mirror along with a detective control to enforce Amazon SageMaker resources being attached to a VPC. The template will also create a product portfolio in Amazon Service Catalog which enables users with appropriate permissions to create a data science environment dedicated to a single project. Once this environment is created you will move on to the next lab which will use Amazon Service Catalog to provision a SageMaker notebook.
Deploy a CloudFormation template which creates a product portfolio, a detective control, and a PyPI mirror:
|Oregon (us-west-2)||Deploy to AWS Oregon|
|Ohio (us-east-2)||Deploy to AWS Ohio|
|N. Virginia (us-east-1)||Deploy to AWS N. Virginia|
|Ireland (eu-west-1)||Deploy to AWS Ireland|
|London (eu-west-2)||Deploy to AWS London|
|Sydney (ap-southeast-2)||Deploy to AWS Sydney|
Visit the CloudFormation console and select the
Outputs tab for the stack you just deployed.
Use the provided link to assume the role of Project Administrator.
Navigate to the Service Catalog console and launch a Data Science Project Environment.
Next. Note S3 bucket names need to be globally unique, so don’t use project-abc verbatim but replace “abc” with something unique such as “yourname-12345” etc.
Nextwithout entering any Tag Options.
Nextwithout checking any Notifications.
The product will take approximately 10 minutes to launch.
You have now created the underlying infrastructure for a secure data science environment which enables project teams to self service. In the next lab you will use this environment to create a Jupyter notebook for yourself.