Hello. I’m a DBA, and I want to understand how Liquibase works as I’m working on a specific case.
To explain :
At work, we have an Oracle database, and the schema modifications are handled by Liquibase on the devs’ side, schema by schema.
But some Oracle objects are not even in the code or in Liquibase, as the database schemas are quite old(really old!) and we have inconsistencies between the current Liquibase changeset and the current Oracle database schema (for the moment, I’m taking only one schema as an example)
My idea, and I don’t know if it’s feasible, is to generate a new base changeset/changelog (excuse me, I don’t know the difference between the two) from the database and start from the new changelog to have a proper Liquibase XML changelog. If I understand correctly what the devs said to me, in order to rerun this new changelog as a base, we will have to delete or truncate the existing tables DATABASECHANGELOG and DATABASECHANGELOGLOCK.
Am I right, and is it the good way to synchronise both Liquibase and the Oracle schema?
Thank you for the help
Hécate