Connecting to a highly available Content Platform Engine

The configuration of the Content Platform Engine connection in a highly available environment depends on the type of application you want to connect to the Content Platform Engine. You can configure administrative applications or user applications to connect to Content Platform Engine.

Administrative applications, IBM® FileNet® Deployment Manager, modify metadata and global configuration database information such as property templates, class and property definitions, and domain level objects such as sites, object stores, and file storage areas.

Due to the built-in lag in metadata synchronization across the Content Platform Engine nodes, changes made on one node are not synchronized immediately after a configuration change is made via an administrative application. To operate properly, the administrative applications, including FileNet Deployment Manager must be configured to use a Content Platform Engine URL that does not get load balanced within the cluster, but instead works with a single Content Platform Engine instance.

In the event the configured Content Platform Engine fails you must reconfigure your administrative applications to point to a different Content Platform Engine node. For details, see Global configuration databases.

For user applications, the Content Platform Engine connection in a highly available environment differs depending on your HA configuration, but in any case must be configured to connect to the virtual server, so that it will be redirected automatically to another Content Platform Engine node in the event of a Content Platform Engine node failure.

• If a Java™ application server cluster is used, EJB connections must use a specific format for establishing connections to Content Platform Engine.
• Applications that use the Web service (CEWS) protocol to connect to Content Platform Engine must use the virtual server name of the hardware or software that is load balancing the connection.