A cluster is a set of WebLogic servers working together, serving pages at the same port. From the user’s point of view, all of the servers function as a single server; it doesn’t matter which server handles a given request.
Virtually all production sites use clustering. Clustering provides much better performance and reliability than running on a single server. For example, if one server in a cluster goes down, the user will not be aware of it, because the other servers in the cluster can take over the sessions it was handling.
Setting up clustering of WebLogic servers running ATG applications involves the following steps:
Create a group of WebLogic servers for serving pages, and assign them to a cluster.
Create additional WebLogic servers for the ATG lock manager, process editor server, workflow process manager and any other services that require a dedicated server. Assign these servers to a different cluster from the page servers.
For each WebLogic server, create a corresponding ATG server configuration.
Assemble your ATG application, and deploy it on each WebLogic server in both clusters. Configure the application on each WebLogic server to use the ATG server configuration that corresponds to that server.
ATG recommends that you use replicated session persistence in your WebLogic clusters, rather than the default memory configuration.
See the Oracle WebLogic documentation for information about creating WebLogic servers and clusters. For information about creating ATG server configurations, see Creating Additional ATG Server Instances in the Configuring Nucleus Components chapter.
