Hi everyone,
I’m starting to work with SQLcl projects and Liquibase (LB) integration for CI/CD pipelines, and I have some questions about best practices when dealing with existing databases.
- Database already exists with many soft and hard objects (tables, views, packages, procedures, triggers, etc.) Many of the schemas in our environment share and use objects across other schemas.
- Database has never used Liquibase before.
- What DB objects should be included moving forward.
Some of my questions:
- Baseline creation – Should I export the entire schemas into repo and use
changelog-sync for Liquibase and make that my baseline is it better to only include selected objects that are still relevant for development going forward?
- Dev → Prod workflow – In good development/production setups, what’s the recommended way to handle existing objects so that SQLcl project can take over smoothly?
Thanks in advance!