Question

A retail company runs a monolithic e-commerce application on-premises backed by a single large PostgreSQL database. Leadership wants to modernize by decomposing the monolith into microservices, each owning its own data store, deployed on Amazon ECS on Fargate. The architecture team is concerned that the current database has heavy cross-domain foreign-key relationships (orders, inventory, customers, and payments are all tightly joined). They need a migration and modernization approach that reduces the risk of a failed cutover while allowing the team to extract services incrementally. Which approach best meets these requirements?

Accepted Answer

Apply the strangler fig pattern: identify bounded contexts within the schema, place a routing/facade layer in front of the monolith, and incrementally extract one domain at a time into a microservice with its own database, using AWS DMS with change data capture to keep data synchronized during each extraction.

Answer

Use AWS MGN to rehost the entire monolith and its database to EC2 first, then rewrite all services simultaneously in a single big-bang cutover once the lift-and-shift is complete.

Answer

Immediately split the PostgreSQL database into four separate Amazon Aurora databases based on table groupings and repoint the monolith at all four, then containerize the application without changing its code.

Answer

Migrate the monolithic database to Amazon DynamoDB using AWS DMS, then rebuild all microservices to share the single DynamoDB table via item-level access patterns to preserve the existing relationships.