Implementing Siebel eBusiness Applications on DB2 UDB for z/OS and OS/390 > Understanding Siebel Partitioning > Strategies for Special Types of Tables >

Partitioning S_ESCL_REQ and S_ESCL_LOG Tables

Tables S_ESCL_REQ and S_ESCL_LOG, which are used by Assignment Manager and Siebel Business Process Designer, tend to grow very large. S_ESCL_REQ is processed based on the GROUP_ID, whereas S_ESCL_LOG is processed by RULE_ID.

Partitioning of these tables can be done in one of the following ways:

• Hard-coding the actual values for GROUP_ID and RULE_ID. This approach is recommended when the number of groups is static. Additional partitions can be defined to accommodate new values that are unknown at creation time. To use this option, you must first define the groups and then extract the actual GROUP_ID values you want to use for partitioning.
• Adding the PARTITION_COLUMN and a trigger to populate the column. The number of partitions defined should ensure that data from multiple groups goes into multiple partitions. Start with 36 partitions, because this is the base number for the Siebel row ID.

NOTE:  If you encounter resource contention due to multiple groups using the same partition when you create new groups and assign new group IDs, consider increasing the number of partitions to allow for a more granular level of GROUP_ID assignments to a partition range.