Which solution should a solutions architect recommend?
Set the backup retention period to 90 days when creating the RDS DB instance.
Configure RDS to copy automated snapshots to a user-managed Amazon S3 bucket with a lifecycle policy set to delete after 90 days.
Create an AWS Backup plan to perform a daily snapshot of the RDS database with the retention set to 90 days. Create an AWS Backup job to schedule the execution of the backup plan daily.
Use a daily scheduled event with Amazon CloudWatch Events to execute a custom AWS Lambda function that makes a copy of the RDS automated snapshot. Purge snapshots older than 90 days.
Explanations:
Aurora’s maximum automated backup retention period is 35 days, so setting it to 90 days is not possible.
RDS does not natively support copying automated snapshots to S3. Additionally, manual lifecycle management of RDS snapshots is complex and not feasible in this scenario.
AWS Backup allows setting custom retention periods beyond the RDS maximum of 35 days. By configuring AWS Backup with a 90-day retention policy, the requirement can be met efficiently.
This approach requires custom Lambda functions for snapshot copying and purging, which is less efficient and more complex to manage than AWS Backup, which has built-in support for 90-day retention.
I believe the answer is:
Create an AWS Backup plan to perform a daily snapshot of the RDS database with the retention set to 90 days. Create an AWS Backup job to schedule the execution of the backup plan daily.