Download now Free registration required
Once a procedure or function is established and is put to use it becomes very difficult to bring any changes in the program. One has to be very careful about how he or she changes the header or signature of the program in use, as any change that one makes in the header is likely to cause a number of other changes throughout the rest of your code base. In simple terms this means that dependent programs will be invalidated and will have to be recompiled. Changes must also be propagated from the development instance, through one or more test instances, and finally into the live production environment. The better one manages the header change-and minimize the impact of that change-the shorter the downtime your production instance will experience as one upgrades their application code base. This paper proposes how Oracle Database helps in making these changes simple and easier. Oracle Database automatically maintains dependency information about all the objects in the database. Whenever one changes a database object, all dependent PL/SQL program units are then marked invalid as they need to be recompiled before they can be used. Thus, Oracle Database does an awful lot of work by keeping our code base in sync. In Oracle Database 11g, Oracle offers fine-grained dependency tracking: the granularity of dependency information within a program unit. This means that Oracle Database will mark a dependent program unit as invalid, thus minimizing the need for recompilation.
- Format: HTML
- Size: 0 KB