Story

Siebel Database Layout Guidelines

As with most client-server applications, the overall performance of Siebel eBusiness applications is largely dependent on the I/O performance on the database server. To ensure optimal I/O performance, it is critical that the tables and indexes in the database be arranged across available disk devices in a manner that evenly distributes the I/O load.

The mechanism for distributing database objects varies by RDBMS, depending on the manner in which storage space is allocated. Most databases have the ability to assign a given object to be created on a specific disk.

A redundant array of independent disks, or RAID, can provide large amounts of I/O throughput and capacity, while appearing to the operating system and RDBMS as a single large disk (or multiple disks, as desired, for manageability). The use of RAID can greatly simplify the database layout process by providing an abstraction layer above the physical disks while ensuring high performance. Regardless of the RDBMS users implement and the chosen disk arrangement, be sure that users properly distribute the following types of database objects:

Database log or archive files.

Temporary workspace used by the database.

Tables and Indexes:

In most implementations, the tables and corresponding indexes in the following list tend to be some of the more heavily used and should be separated across devices. Refer to the Siebel Installation Guide in Siebel Bookshelf for a more complete listing. In general, the indexes listed below should be on different physical devices from the tables on which they are created.

o S_ACCNT_POSTN
o S_OPTY
o S_ADDR_ORG
o S_OPTY_POSTN
o S_CONTACT
o S_POSTN_CON
o S_DOCK_TXN_LOG
o S_PARTY_REL
o S_PARTY
o S_SRV_REQ
o S_EVT_ACT
o S_OPTY
o S_ORG_EXT

For Organizations who plan on making extensive use of Siebel Enterprise Integration Manager (EIM), it is recommended to put the key EIM tables (based on the unique business requirements) and their corresponding indexes on different devices from the Siebel base tables and indexes, because all are accessed simultaneously during EIM operations.