The operation of removing redundant data from one table into another table is called table normalization. Table normalization should be used to improve the storage efficiency of table data as well as to enhance the capacity for concurrent executions during access processing. By normalizing a given table repeatedly, it becomes possible to transform a complex database into an optimal database. Figure 3-5 illustrates table normalization. For details about table normalization, see the HiRDB Version 8 Installation and Design Guide.

Figure 3-5 Table normalization

Explanation

Before normalization, the columns PCODE and PNAME in the STOCK table have a 1-to-1 correspondence; the information in these two columns is redundant. In such a case, you can create another table, named PRODUCT in this example, composed of the PCODE and PNAME columns from the STOCK table, so that in the PRODUCT table there is no duplication of information in the PCODE and PNAME columns.