When you are into database designing, you may consider normalization as the only method of designing. But there can be several other designing rules. Here are glimpses of few such rules which you are required to keep in mind while designing the database.
Identify The Nature Of Application: At first, you are required to identify the type of the application, on which you are designing. You should detect whether the application is analytical or transactional in nature.
- Analytical: Here, the end users will be interested in reporting, forecasting and the analytical functions. These databases will possess small number of updates as well as inserts. OLAP can be considered as the official name for this analytical application.
- Transactional: In this type of application, the end users will involve in the CRUD functions i.e., create, read, update and delete the records. OLTP can be the official name for the transactional application.
Target All The Non-Uniform Duplicate Data As Enemy: Duplicate data should not fill up spaces in your hard disc. You can move data from one table to a unique master table and can refer through the foreign keys.
Partial Dependents: There can be various fields which are partially dependent to each other. For example, if you are placing the data of the syllabus in the primary keys, the standard and the section are the partially dependent fields.
Centralize The Table Design Of Name And Value: With the Oracle remote dba, you will be able to centralize the design for the name and the values. The tables of the names and the values will include the keys and the data associated with those keys.
Performance Should Be The Key: You cannot strictly avoid redundancy all the time, if performance remains your key. You can think of de-normalization activities if there are requirements in your performance. You may require creating various joints in the normalization process and the tables are to be created within the de-normalization process.
Multidimensional Data: The multidimensional data are generally dealt with the OLAP projects. For example, if you are recording data of sales in each country, you can include fields like the name of the organization, date and the sales in each country. The sales figures will therefore possess data of various dimensions. You can create the tables for the central sales and create a connection with the tables through the relationship of the foreign keys.
Hierarchical Data: You can include unlimited data through hierarchy method. For example, if you are presenting a marketing scenario with various sales persons under them, you can surely use the hierarchical method. Foreign keys and self referencing primary keys will be helpful in obtaining your goal.
So besides normalization techniques, you can use all the above mentioned rules for database designing. You should always select the derived columns wisely and select the keys accordingly. Create tables in such a way that they remain multidimensional and interdependent to each other. So the referencing functions will be more purposeful with the tables.