Hello, I'm in the process of converting a Rail application over to Rails 6 from Rails 5. In that application, we have a large number of databases and we do some parallelization to speed up creation, migration, and test preparation for those databases.
Previously, we were using multiverse <https://github.com/ankane/multiverse> for this. When we ran a top-level bundle exec rails db:migrate, the database would be dumped to the schema file. Today in Rails 6, that's still the case when running all migrations <https://github.com/rails/rails/blob/1605a631fb7b590117604d0208a461e55aa3df2e/activerecord/lib/active_record/railties/databases.rake#L82-L91>, but when running database-specific migrations with bundle exec rails db:migrate:DATABASE, the schema is not dumped <https://github.com/rails/rails/blob/1605a631fb7b590117604d0208a461e55aa3df2e/activerecord/lib/active_record/railties/databases.rake#L109-L116> . Unfortunately, that means that we can only parallelize the creation and migration of the databases. After that, we have to serially run bundle exec rails db:structure:dump and bundle exec rails db:test:prepare. I'd like to propose supporting database-specific tasks for db:structure:dump and db:test:prepare to handle this situation. I'm willing to open PRs to support this assuming this seems reasonable to everyone. Thanks! - Kyle Thompson -- You received this message because you are subscribed to the Google Groups "Ruby on Rails: Core" group. To unsubscribe from this group and stop receiving emails from it, send an email to rubyonrails-core+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/rubyonrails-core/f41aabaa-c381-4783-bb5c-2de78f7ad043%40googlegroups.com.
