Database releases. How can we make things better?

Releases of Java (technology related) code often are easy. Releases of database code often are not. Why is that? An important difference and cause can be found in that database code tends to be incremental. Being incremental means that a database release is build on top of previous releases. A database release has to be incremental because often a production environment holds data which needs to be preserved. With Java code, the previous releases (EAR, WAR, JAR files and the like) are less important for the present release since usually they are overwritten. Of course Java releases have their own complexities, however often they do not hold data; that is the task of the database. In this blog I will describe some of the common issues and several suggestions to improve (database) releases. Even though I am a developer (with quite some experience in release automation), I tried to also describe the issues system operators are facing and provide suggestions to improve the process.