Siebel Analytics Performance Tuning Guide > Performance Tuning >

Overview of Performance Tuning

Siebel Analytics allows users to access data distributed across multiple independent databases through a single, unified, and easily understood business model. Therefore, when you tune Siebel Analytics for optimal performance, you must also tune the underlying database servers, the communication network, and the extract, transform, and load routines. Otherwise, performance will fall below optimal.

You must also make sure that sufficient hardware resources are available to support the entire workload. That means, a sufficient amount of memory, a sufficient number of processors, a sufficient amount of physical disk space with the requisite characteristics, and ample network bandwidth to support the entire workload.

This chapter summarizes specific actions you can take to improve the performance of Siebel Analytics, the underlying database servers, and the extract, transform, and load routines. The actions are organized into three areas:

Subsequent chapters describe these actions in more detail and provide examples.

All the actions described represent steps which have proven effective in a large number of cases, and most database administrators and performance analysts consider these actions as part of their standard practices.

Keep in mind that these actions are intended as guidelines, and the results depend highly on your particular workload, data shape, and requirements. Before summarizing the performance actions, the next section reviews a few working assumptions as well as a suggested approach to performance tuning. You should also review the logical and physical database schemas from a performance point of view. Figure 1 shows the three performance areas addressed in this guide.

Figure 1. The Three Areas

Click for full size image


 Siebel Analytics Performance Tuning Guide 
 Published: 18 April 2003