Netscape Application Server-- Balancing User-Request Loads

Administration Guide
Overview Guide
Installation Guide


About Load Balancing
Load balancing benefits the user because their requests, not having to wait for a busy server to become available, are returned to them more quickly. For example, many users are accessing an application during a peak usage hour. Most of the application's AppLogic objects are duplicated on several Netscape Application Servers and enabled for load balancing. As one Netscape Application Server reaches its optimal handling capacity, subsequent requests are distributed to another Netscape Application Server. With multiple servers handling all the requests, the user does not have to wait an unacceptable length of time for results. How Load Balancing Works The distribution of requests across multiple Netscape Application Servers is monitored and assigned by a load balancing service built into each Netscape Application Server. The load balancing service determines which Netscape Application Server is best-suited to handle a request by comparing servers based on the following two values: The Server Load value The Server Load value is calculated from various system criteria, such as the CPU load and amount of memory thrash. These criteria are indicative of how busy a Netscape Application Server is and, therefore, how well it might handle additional requests. For more information about these values, see "About the Server Load Value." The AppLogic Performance value The AppLogic Performance value is calculated from various AppLogic criteria, such as the number of times the object was run on a particular server or whether the AppLogic results were cached. These criteria are indicative of how well the AppLogic object is running on a particular Netscape Application Server. For more information about these values, see "About the AppLogic Performance Value."