Based on the availability requirements for the Dynamic Service Activator, you can set up Dynamic Service Activator to run in a redundant environment in the following ways:
The Web application server on each system may already be installed in an environment that uses load-balancing software.
Each SOAP request tries to activate or deactivate the same service session at the same time. This scenario gives better subscriber response time than spaced requests from the service.
For a configuration in which services send SOAP requests to two Dynamic Service Activators, you can optimize the code to wait only for the first call. As soon as one of the SOAP requests is completed, the service application can continue and not wait for the second call to be completed. To detect a failure of the Web Services Gateway or the client’s network connection, the application needs to wait for both calls to time out before the request fails.
Duplicate requests and responses places a higher load on the system that runs the service application and on the system running the SAE.
