Storing Data Efficiently
Specifying DECIMAL and NUMERIC Precisions
Storage requirements for decimal data types are a function of the precision required of the values to be stored. The following table indicates the number of bytes used to store DECIMAL and NUMERIC data types of various precisions.
|
Number of Precision Digits |
Number of Bytes Stored |
|
1 - 2 |
1 |
|
3 - 4 |
2 |
|
5 - 9 |
4 |
|
10 - 18 |
8 |
You can achieve optimal space savings by using multivalue compression together with an efficient decimal size. An example would be to define decimal precision such that a 4-byte representation was used instead of an 8-byte representation, for example DECIMAL(9) instead of DECIMAL(10). Then, multivalue compression could be used on the most frequently occurring values to reduce the storage overhead still further.