Example of a Partition Definition
If you use the table S_ADDR_ORG, the column OU_ID appears to be a good candidate for a partitioning key. However, OU_ID contains data in the Siebel row ID format and row IDs are generated in ascending order. Instead a trigger is used to generate a random number between 00 and 10 and this number is stored in a new physical PARTITION_COLUMN. This column is defined in the storage control file for S_ADDR_ORG.
Note: If you have evenly sized table partitions, using PARTITION_COLUMN
as the partitioning key provides query performance optimization. If
you do not have evenly sized table partitions, a different partitioning
column might be more effective in enhancing performance and scalability.
For tables that use table-controlled partitioning, you must define two types of objects in the storage control file: PartitionBase and PartitionPart. These objects specify the partitioning key column and partition value ranges. An example definition for each is shown.
Note: If you are defining a partitioning index, you must create IndexBase
and IndexPart objects to specify the partitioning key and value ranges.