Which solution will meet these requirements with the LEAST operational overhead?
Use Amazon DynamoDB with connection pooling with a target group configuration for the database. Change the applications to use the DynamoDB endpoint.
Use Amazon RDS Proxy with a target group for the database. Change the applications to use the RDS Proxy endpoint.
Use a custom proxy that runs on Amazon EC2 as an intermediary to the database. Change the applications to use the custom proxy endpoint.
Use an AWS Lambda function to provide connection pooling with a target group configuration for the database. Change the applications to use the Lambda function.
Explanations:
Amazon DynamoDB is a NoSQL database service, which would require a significant change to the applications to use a different database endpoint and structure. This does not address the requirement of scaling the existing RDS database effectively.
Amazon RDS Proxy provides connection pooling and can handle scaling of the database connections automatically, minimizing operational overhead. This option allows applications to seamlessly connect to the database while managing connections efficiently.
A custom proxy running on Amazon EC2 would require additional management and operational overhead, such as maintaining the EC2 instances, scaling, and ensuring high availability, which goes against the requirement of minimizing operational overhead.
Using an AWS Lambda function for connection pooling may not be effective for handling persistent database connections. Lambda is designed for stateless, short-duration tasks, which can lead to latency issues and does not inherently provide connection pooling capabilities like RDS Proxy.
It appears that the answer is:
Use Amazon RDS Proxy with a target group for the database. Change the applications to use the RDS Proxy endpoint.