Stairway to Database Source Control
The goal of this Stairway series is to provide a brief, practical and systematic guide to source controlling your SQL Server databases, so that you can then manage your ongoing development and deployment from a known version, within source control.
- The first level of this stairway introduces the basics of source control, some common technologies and technologies, and demonstrates how to start versioning a SQL Server database.
- In this level, we're going to continue the philosophy of learning by example, and get a database into our SVN repository. We will also consider our overall approach to source control for databases, and the manner in which our team will develop these databases, concurrently.
- One of the main purposes of placing a database under source control, alongside the application code, is to allow team collaboration during development projects. The Version Control System (VCS) stores and manages all of the project files, maintaining an audit trail of what changed, and who made the change. Each team member can work on a file, or set of files, and submit their changes to the VCS to make them available to other team members. They can also inspect the VCS to discover recent changes made by other team members.
- Now that we have our database under source control, we will want to share our work with other developers. If we are in a centralized source control system, our changes may be committed straight into the central repository.
When we are working in a distributed system, it means pulling down any changes from other developers, addressing any areas of conflict, and pushing our changes up to allow others to benefit from our work. This allows our changes to be synchronized with the changes other developers have made.
This level is principally about setting up a distributed source control system, namely Git, and how to commit database development changes to a local repository, before pushing them into a remote 'central' repository for sharing with other developers.
The next level will delve a little deeper into Git's versioning mechanisms, and show some examples of how to share database changes during development, and how to deal with conflicting changes.
- This level starts with an overview of how versioning works in Git, a DVCS, and suggests a sensible database project versioning strategy. It then offers some simple, but illustrative, practical examples showing how to share database changes and deal gracefully with any conflicting changes.