Cluster topology

Figure 1 shows a cluster topology containing two servers on which separate instances of the FTM SWIFT enterprise applications run, and a third server on which the administrative console runs.

The two instances of the FTM SWIFT enterprise applications are both connected to a single http server on which the WebSphere® plug-in is installed. That plug-in distributes the workload between the two application servers by routing requests to each FTM SWIFT enterprise application in turn. If one application server becomes unavailable, the plug-in recognizes this and routes subsequent requests only to the cluster member that is still available. After the failed application server becomes available again, the plug-in resumes routing requests to both application servers. This scenario is often used in production environments that require high availability, workload balancing, and failover capability.

The application servers on HostA and HostB are members of the same application server cluster. The configurations of each of the FTM SWIFT enterprise applications is administered centrally by means of the deployment manager (dmgr) on HostC. The separate network deployment manager and node agent employed by a cluster topology require additional resources.