In our application, we use the same entity model with the first approach to the database, and we have one db scheme on the server. We also use git to control version control.
Problems arise when some of our developers work on new features that require changes to the database, and this happens often, so our solution at the moment is as follows:
- Create a new branch from the main branch.
- Create a new db for testing that matches the one we use in it.
- Having finished transferring the new functionality to the main branch and pulled it out of the newly created one, take a new entity model with a new db change and apply sql on our live db when publishing.
Problems arise if we stop for too long on some new branch, because we often change industries.
At that time, our main branch changed a lot, as well as live db with it, so our db test structure is very different from the time when we created a new branch, and then we had problems combining our new branch with master one.
We do this successfully every time, but it’s difficult to deal with. I want to know if someone knows the best management system and workflow.
Aleks source
share