Network Load Balancing

Previous Topic Next Topic

How Network Load Balancing Works

Briefly, when Network Load Balancing is installed as a network driver on each of the cluster hosts, the cluster presents a virtual IP address to client requests. The client requests go to all the hosts in the cluster, but only the host to which a given client request is mapped accepts and handles the request. All the other hosts drop the request. Depending on configuration of port rules and on affinity, the statistical mapping algorithm, which is present on all the cluster hosts, maps the client requests to particular hosts for processing.

The hosts exchange heartbeat messages to maintain consistent data about the cluster's membership. If a host fails to send or does not respond to heartbeat messages, the remaining hosts perform convergence, a process in which they determine which hosts are still active members of the cluster. If a new host attempts to join the cluster, it sends heartbeat messages that trigger convergence. After all cluster hosts agree on the current cluster membership, the client load is repartitioned, and convergence completes.

Discussion of Network Load Balancing clusters requires clarification of two kinds of client states, application data state and session state:

By setting port rules, cluster parameters, and host parameters, you gain great flexibility in configuring the cluster, which enables you to customize the cluster according to the various hosts' capacities and sources of client requests. You can:

You can combine the preceding capabilities by setting cluster and host parameters and creating port rules for your particular scenario. For guidelines on setting parameters and port rules for various scenarios, see "Scenarios" later in this chapter.

Before specific scenarios are discussed, the following sections explore the basic concepts of Network Load Balancing:

© 1985-2000 Microsoft Corporation. All rights reserved.