Should we choose "Lift and Shift" or "Refactor" for our first major move to the cloud?

All answers to this question.

This is the classic "Migration Dilemma." If you have a hard deadline (like a data center lease ending), Lift and Shift is your only real choice. It gets you there fast, but you won't see the cost benefits of the cloud until you start using managed services. However, if you have the time, "Replatforming" is a great middle ground. This involves moving the app as-is but switching the database to a managed service like RDS. In 2023, we saw that companies who refactored even 20% of their legacy code during the move saved significantly more in the long run than those who just moved VMs.

   Answered 2025-03-21 by Alice Vance

What is the specific business goal? Is it to save money immediately, or to gain the ability to release software features faster?

   Answered 2025-03-23 by Philip Duncan

• Philip hits the nail on the head. If the goal is agility, you have to refactor. If you just "Lift and Shift" a monolith into , your deployment cycles will still be slow and painful. You'll be paying for the cloud's elasticity but won't be able to use it because your app isn't designed to scale horizontally. We learned this the hard way in 2024 when we moved a legacy system and our "Agility" metrics didn't move an inch until we broke the monolith into smaller microservices months later.

     Commented 2025-03-24 by Roy Chapman

Do a "Pilot" migration with one small, non-critical module first. It will reveal all the hidden dependencies you didn't know your ERP had.

   Answered 2025-03-25 by Janet Peterson

• The "Pilot" approach saved our skin. We found a hardcoded IP address in a legacy script that would have broken everything if we moved all at once.

     Commented 2025-03-26 by Joyce Gardner