Question

A financial services company runs a monolithic COBOL batch processing application on an aging on-premises mainframe. The application handles nightly transaction reconciliation and represents a significant licensing and maintenance cost. Leadership wants to eliminate the mainframe within 18 months while minimizing the risk of a big-bang cutover, gradually replacing functionality module by module, and adopting modern managed services. Business logic must be preserved but the team wants to move toward event-driven, serverless components over time. Which migration and modernization approach best meets these requirements?

Accepted Answer

Use AWS Mainframe Modernization to automatically refactor the COBOL code to Java, then apply a strangler fig pattern—incrementally routing individual reconciliation functions to new serverless components while the legacy modules continue running until each is replaced.

Answer

Rehost the entire mainframe workload on a single large EC2 instance using an emulator, then leave it unchanged since business logic must be preserved.

Answer

Perform a complete rewrite of the application into AWS Lambda functions in a single coordinated release, cutting over all batch jobs at once at the end of the 18-month window.

Answer

Retain the mainframe on premises and use AWS DMS to replicate transaction data to Amazon Redshift for reporting, deferring the modernization decision indefinitely.