6.3 Web container clustering and failover (Web server plugin)

Each HTTP server is configured to run the IBM HTTP Server Plug-in. The cluster members can all reside on a single node or can be distributed across multiple nodes in the WebSphere cell (vertical or horizontal scaling).

Each request coming into the Web server is passed through the plug-in, which uses its configuration information to determine whether the request should be routed to WebSphere, and if so, to which application server (that is, to which Web container) the request should be routed to (Figure 6-1). The communication between the plug-in and the application servers can be either HTTP or HTTPS. The Web server plug-in distributes requests around cluster members that are not available.

Figure 6-1 Plug-in (Web container) workload management

The plug-in is a module, which is loaded by the Web server when it initializes, and that it bypasses Web containers that are not available. It uses the following mechanisms for WLM and failover:

• Application server clustering, which creates server process redundancy for failover support. All application servers in a cluster host the same applications.

• The workload management routing technique built into the IBM HTTP Server Plug-in. It controls the routing of client requests among redundant server processes. This routing is based purely on the weights associated with the cluster members. If all cluster members have identical weights, the plug-in sends an equal number of requests to all members of the cluster, when assuming no session affinity. If the weights are different, the plug-in routes requests to those cluster members with the higher weight value more often.

• Session management and failover mechanism, which provides HTTP session data for redundant server processes.

Thus, satisfactory failover support for Web clients can only be achieved by the use of all three mechanisms.

xxxx