The Parlay specification does not define the process of locating the Parlay gateway, but it does provide service type information and interface definitions to be used when configuring the gateway location.
The Parlay Connector uses configuration data to control dynamic loading of the gateway location class. You can configure the connector to accept the default gateway location class or to use a customized gateway location class. You can configure the default gateway location class for some customization during the location process. If a customized gateway location class is used, it will be passed a hash table containing the Parlay Connector configuration and state information.
For failover, you can configure a secondary Parlay gateway as a backup in case the primary gateway fails. For each additional gateway, you must configure unique service IDs.
The Parlay-based Web service implementations use the gateway session IDs and assignment IDs as references to track requests sent to the gateway.
Note that if the primary gateway fails and the secondary, or backup, gateway assumes the workload, the session IDs and assignment IDs for each gateway must be unique or an exception is generated.