I’ve been considering an approach to development that could streamline the migration process to a test server, especially when dealing with a large number of files or significant file sizes that typically result in long migration times.
In a typical migration, 25-30 files are transferred, including a few interface files that interact with data files. Given this structure, could a developer proactively track schema changes—such as new fields or other additive modifications—to ensure that migrating a data file independently wouldn’t impact production?
The idea is that a full migration can take over four hours, but if we can establish high confidence that all schema changes to data files are strictly additive, we could implement smaller, scheduled deployments of these data files throughout the week without disrupting testing or production. Then, during a scheduled full migration, only the interface files or other files with interdependent code changes would need to be migrated simultaneously.
Would this kind of incremental approach improve efficiency while minimizing downtime?
I open the floor for sage banter.