Database Deployment: Synchronizing Database Structures

The simplest and most obvious way of synchronizing two databases would be by deleting the old database and creating new structures on top of it. If you back up the data before deleting the old database, and restore it from the backup after you're finished updating the structures, you might just get away with it. However, this approach is unacceptable in many environments as it disrupts the workflow and does not provide reasonable security.

Coding a set of SQL statements and running the script on the client side is a better idea, but still has many weaknesses as it consumes a developer's time on routine operations, and opens the possibility for errors. If you are upgrading a database to a new version, you'd have to apply all intermediate patches, introducing a lot more possibilities for something to go wrong.

Thinking logically, one can purchase a specialized synchronization tool, install it on the developer side as well as distribute it among the customers. This does allow a certain time-saving, but the method is costly and often requires paying for client-side licenses.

If you don't want to code all of the updates manually, and don't want to distribute third-party software to perform updates with your application, consider implementing the necessary update functionality into your own application. The built-in functionality will handle structural updates and database upgrades directly from within your application.