In this section, we summarize the capabilities of the WebSphere Load Balancer, which is part of the IBM WAS ND V6 Edge Components, and provide an overview of how to set up a highly available Load Balancer pair.
The Load Balancer has several components that can be used separately or together to provide failover and high availability support:
We are mainly interested in the Dispatcher component which can be used to distribute workload between multiple Web servers (that is, a Web server cluster). This provides both, scalability and high availability for the Web servers. The workload is dispatched to the Web servers in the cluster based on a load balancing mechanism usually known as IP spraying, which intercepts the HTTP requests and redirects them to the appropriate machine in the cluster, based on weights and availability.
Users must be able to reach the application regardless of failed servers. In a clustered Web server environment, the Load Balancer monitors the availability of the Web servers. If a Web server has failed, no more requests are sent to it. Instead, all requests are routed to the remaining active Web servers. See Web server for additional information.
Being the entry point into your WebSphere system, it is extremely important that your Load Balancer is highly available. This can be achieved by setting up a second Load Balancer server with exactly the same configuration. You have to define one or more heartbeat connections between these servers. The heartbeat mechanism is built into the WebSphere Load Balancer. You have two options to configure such a highly available Load Balancer environment:
For details on Load Balancer configuration, see Chapter 4 and Chapter 5 of the IBM WebSphere V6 Scalability and Performance Handbook, SG24-6392, where step-by-step configuration instructions are provided.
Figure 15-2 shows an Active/Passive configuration before a failover to the backup LB system.
Figure 15-3 shows the takeover of the cluster IP address by the backup Load Balancer server in case of a failure of the primary LB server.
The Load Balancer can also be used to provide high availability for other components, such as firewalls and directory servers. This is covered in the various components' sections in this chapter.
