Ah, I think I understand better. I can't answer for any kind of specific enterprise environment, but I'm sure there are many different systems cobbled together with all sorts of creative baler twine and you could get a wide variety of answers to this question.
I'm actually working on a project right now where we're trying to keep data updated between two systems. The incoming data gets imported to a MySQL database and every now and then, new data is exported to a .sql file. Each row has an auto incrementing primary key "id", so we very simply keep track of the last exported ID and start the export from there (using mysqldump and the --where argument). This is quite simple and doesn't feel like an "enterprise-ready" solution, but it's fine for our needs. That avoids the problem of duplicated inserts.
Another solution would be to export the entire database from your development system, then through some series of actions import it to the production server while deleting the old database entirely. This could depend greatly on the size of your data and how much downtime you're willing to perform. An efficient and robust implementation of this would import to a staging database (and verify there were no import problems) before moving the tables to the "correct" database.
If you are simply referring to schema changes or very small amounts of data, then probably version control is your best bet. This is what I do for some of my database schemas; basically you start out with the base schema, then any change gets written as a script that can be run incrementally. So for instance, in an inventory system I might have originally started with a customer table, with fields for ID and name. Later I added a marketing department, and they want me to get email addresses. 2-email.sql would be this line: ALTER TABLE `customer` ADD `email` VARCHAR(255) NOT NULL AFTER `name`;. Still later, if I decide to handle shipping, I'll need to add mailing addresses, so 3-address.sql adds that to the database. Then on the other end, I just run those through a script (bonus points are awarded for using MySQL logic such as "IF NOT EXISTS" so the script can run as many times as needed without error).
Finally, you might benefit from setting up a replication system. Your staging database would automatically send all changes to the production database. Depending on your development process, this can be quite useful or might just get in the way.
