Implementing Siebel Business Applications on DB2 UDB for z/OS > Understanding Siebel Partitioning >

Considerations in Partitioning Tables

Careful planning, requirements analysis, and monitoring are necessary to achieve optimal partitioning. Decide which Siebel tables to partition and how to partition them on the basis of table size and usage in your deployment. You can partition any table in accordance with your business requirements; you can select all the tables that Siebel Systems provides for partitioning, or you can select a subset of them.

Choosing a Partitioning Key Column

You can use any column as the partitioning key for a Siebel table. In deciding which column is most appropriate for partitioning, consider data access and data distribution factors. You also need to determine the number of partitions your implementation requires.

NOTE:  When choosing a partitioning key for a Siebel table, follow all the rules, restrictions, and concerns listed in the IBM DB2 documentation. Factors to consider include data distribution and potential updates of partitioning keys.
For example, when partitioning keys are updated, system performance is adversely affected. When choosing partitioning keys, you can avoid performance problems by selecting fields that are seldom or never updated.

Choosing a Partitioning Index

The partitioning index is by definition the clustering index in DB2. Therefore, by optimizing your clustering order, you optimize performance. Make sure your partitioning strategy supports more than day-to-day online access to the Siebel tables. Consider processes such as EIM, Siebel Remote, Assignment Manager, and Workflow. See Partitioning Strategies for Special Types of Tables.

Implementing Siebel Business Applications on DB2 UDB for z/OS