Approaches
Replatform
Changing the platform a system runs on, such as its database hosting, runtime version or deployment model, while the business logic stays as it is. It sits between moving a system as-is and rebuilding it.
Part of Maintenance & Modernization
Replatforming changes the platform under an application while its business logic stays the same. Typical examples are moving to a managed database, containerising the application, upgrading a runtime, or replacing a self-managed component with a cloud service.
Where it fits
It sits between a rehost and a rewrite, and often follows a rehost: move first to meet the infrastructure deadline, then replatform to reduce the operational work. Business rules are not rewritten, so the risk of regressions is smaller than in a rewrite.
What changes for the team
A managed database service takes over minor version patching and automated backups, and offers standby replicas for failover. The team still chooses the settings, plans major version upgrades, tests restores and manages access.
Containerising makes deployments repeatable and environments consistent. It can need small code changes, such as reading configuration from environment variables, writing logs to standard output, and keeping files that must survive a restart off the local filesystem.
What causes unexpected effort
Applications often depend on their platform in undocumented ways: a database extension, a file path on the server, a scheduled job on the host. These tend to surface during the move, so we start by listing everything the application touches outside its own code and estimate from that list.
Related
Application Modernization, Cloud Migration & Architecture, DevOps & CI/CD