Which solution will meet these requirements with the FEWEST changes to the architecture?
Migrate the application to Amazon Elastic Container Service (Amazon ECS) containers that use the Fargate launch type in three Availability Zones. Use Amazon S3 to provide file storage for all three containers. Use a Network Load Balancer to direct traffic to the containers.
Migrate the application to Amazon EC2 instances in three Availability Zones. Use Amazon Elastic File System (Amazon EFS) for file storage. Mount the file storage on all three EC2 instances. Use an Application Load Balancer to direct traffic to the EC2 instances.
Migrate the application to Amazon Elastic Kubernetes Service (Amazon EKS) containers that use the Fargate launch type in three Availability Zones. Use Amazon FSx for Lustre to provide file storage for all three containers. Use a Network Load Balancer to direct traffic to the containers.
Migrate the application to Amazon EC2 instances in three AWS Regions. Use Amazon Elastic Block Store (Amazon EBS) for file storage. Enable Cross-Region Replication (CRR) for all three EC2 instances. Use an Application Load Balancer to direct traffic to the EC2 instances.
Explanations:
While Amazon ECS with Fargate offers scalability, it introduces containerization that requires significant changes to the application architecture. Additionally, using Amazon S3 for file storage may not provide the same file system interface needed by the application, potentially complicating the migration.
Migrating the application to Amazon EC2 instances in three Availability Zones provides a straightforward lift-and-shift approach. Using Amazon EFS allows all EC2 instances to share the same file storage, ensuring high availability and redundancy with minimal changes to the architecture. An Application Load Balancer can manage HTTP traffic effectively.
Similar to option A, migrating to Amazon EKS with Fargate introduces containerization, which necessitates significant changes to the existing architecture. Additionally, Amazon FSx for Lustre is more complex and may not fit the application’s storage needs as simply as EFS. A Network Load Balancer may not be ideal for HTTP traffic compared to an Application Load Balancer.
Migrating to three different AWS Regions adds unnecessary complexity and potential latency issues. Using Amazon EBS is also problematic, as it is not a shared file system; thus, it would not work well for multiple EC2 instances without additional complexity for replication. This option does not align with the need for high availability and requires extensive changes.
I deduce that the answer is:
Migrate the application to Amazon EC2 instances in three Availability Zones. Use Amazon Elastic File System (Amazon EFS) for file storage. Mount the file storage on all three EC2 instances. Use an Application Load Balancer to direct traffic to the EC2 instances.