Which solution will meet the company’s requirements?
Configure AWS Backup to perform cross-Region backups of all servers every 5 minutes. Reprovision the three tiers in the DR Region from the backups using AWS CloudFormation in the event of a disaster.
Maintain another running copy of the web and application server stack in the DR Region using AWS CloudFormation drift detection. Configure cross-Region snapshots of the DB instance to the DR Region every 5 minutes. In the event of a disaster, restore the DB instance using the snapshot in the DR Region.
Use Amazon EC2 Image Builder to create and copy AMIs of the web and application server to both the primary and DR Regions. Create a cross-Region read replica of the DB instance in the DR Region. In the event of a disaster, promote the read replica to become the master and reprovision the servers with AWS CloudFormation using the AMIs.
Create AMIs of the web and application servers in the DR Region. Use scheduled AWS Glue jobs to synchronize the DB instance with another DB instance in the DR Region. In the event of a disaster, switch to the DB instance in the DR Region and reprovision the servers with AWS CloudFormation using the AMIs.
Explanations:
AWS Backup does not support a 5-minute backup frequency for applications such as RDS and EC2 instances. Cross-Region backup using AWS Backup is not suitable for meeting the RPO of 5 minutes, and CloudFormation alone cannot automate the full reprovisioning process within such a short recovery window.
Cross-Region snapshots of RDS do not have a 5-minute interval, and snapshots can take longer than the 5-minute RPO requirement to complete. The method described also does not ensure that the application servers are restored quickly enough.
Using EC2 Image Builder to create and copy AMIs of the web and application servers ensures quick provisioning in the DR region. The cross-Region read replica of the DB instance provides near real-time replication and can be promoted in the event of a disaster, meeting the RPO requirement.
AWS Glue jobs are not designed for real-time or near-real-time synchronization of a database. They are batch-oriented, making them unsuitable for maintaining a 5-minute RPO for a live database. Additionally, creating AMIs of the servers in the DR region is not enough to meet the rapid recovery requirements.
I systematize that the answer is:
Use Amazon EC2 Image Builder to create and copy AMIs of the web and application server to both the primary and DR Regions. Create a cross-Region read replica of the DB instance in the DR Region. In the event of a disaster, promote the read replica to become the master and reprovision the servers with AWS CloudFormation using the AMIs.