Planning a CA SiteMinder WSS deployment with performance in mind is the first step to maintaining high enterprise availability and performance standards. A good approach is to estimate the number of expected requests CA SiteMinder® must handle per web service. The following are the most significant factors that influence CA SiteMinder® performance:
Note: Each web service request triggers one authentication and one authorization event.
The following graphic illustrates how request rates fluctuate throughout the day, are sustained for a specific period, and peak within that period:
Note: Authenticating and authorizing requests results in a number of reads from a user store. Determining sustained and peak rates helps you determine the load under which your user stores must operate to service Policy Server requests.
The purpose of the following use case is to illustrate how example.com, a fulfillment service organization approaches capacity planning by modeling the usage of their order fulfillment web service. The use case is referenced throughout this chapter for examples.
The company is planning to deploy CA SiteMinder WSS to protect its web service. The company has 10,000 users in a single user store.
Some web service clients send a single status request to the inventory operation of the web service once a day while others may send as many as four batched order requests per day to the fulfillment operation.
Copyright © 2015 CA Technologies.
All rights reserved.
|
|