The database should be designed to allow features to be less vulnerable to changes by using views and table functions to provide interfaces between features. Design in resilience to change – the team must discuss all planned changes to the database and get an understanding of the potential problems of breaking dependencies.Apply strict database versioning – a great deal of confusion in branching, merging and testing is avoided if it is always clear what version a database is at, the version that any migration script runs from, and the version it migrates to.My general advice for merging database changes as painlessly as possible is as follows: The second is when new features are scheduled for release. The first of these is when changes to the database have to be merged into the development branch. There are several points in a database development where any shortcomings in the methodology in use will become painful. In this article, I’ll offer some advice on how to achieve this. For a database, the objective of each merge is to safely update a database that contains data, without breaking any of the complex interdependencies between its components. Conversely, you might discover that since you created branch, another developer has merged in a change that creates a new dependency on a table that you’ve now altered.Īlmost every engineering process that involves parallel tasks has the same potential problem when it comes to the point of assembly. However, you will sometimes find that a table or view on which your code depends has been altered by someone else in a way that breaks your code. ![]() In a well-disciplined team, merging database changes from a branch back into development will generally proceed with peace, love and understanding. In particular, the extent of the interdependencies can come as a shock. He is a regular contributor to Simple Talk and SQLServerCentral.īranching and merging in database development is similar to branching and merging in software development, but it is a mistake to expect them to be the same. ![]() Phil Factor (real name withheld to protect the guilty), aka Database Mole, has 30 years of experience with database-intensive applications.ĭespite having once been shouted at by a furious Bill Gates at an exhibition in the early 1980s, he has remained resolutely anonymous throughout his career.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |