Which solution should a SysOps Administrator use to allow the Data Scientists to deploy their workloads with MINIMAL effort?
Create an Amazon Machine Image (AMI) of the EC2 instance. Share the AMI with authorized accounts owned by the company. Allow the Data Scientists to create EC2 instances with this AMI.
Distribute an AWS CloudFormation template containing the EC2 instance configuration to the Data Scientists from an Amazon S3 bucket. Set the S3 template object to be readable from the AWS Organizations orgId.
Publish the instance configuration to the Private Marketplace. Share the Private Marketplace with the company’s AWS accounts. Allow the Data Scientists to subscribe and launch the product from the Private Marketplace.
Upload an AWS CloudFormation template to AWS Service Catalog. Allow the Data Scientists to provision and deprovision products from the company’s AWS Service Catalog portfolio.
Explanations:
Sharing an AMI only allows Data Scientists to launch EC2 instances but does not automate or simplify the configuration of the instances, which goes against the requirement for minimal effort.
While distributing a CloudFormation template via S3 would provide the ability to launch instances, it still requires the Data Scientists to manually execute the template, which is more effort than necessary.
The Private Marketplace would allow sharing EC2 configurations, but it focuses on software products rather than instance provisioning. This solution is not the best for simplifying instance creation for the Data Scientists.
AWS Service Catalog allows the SysOps Administrator to define and manage a catalog of approved resources. Data Scientists can then provision and deprovision EC2 instances with minimal effort using pre-configured products in the catalog. This approach meets the need for minimal configuration and maintains control over the instance settings.