Using a J2EE Cluster in the XI 3.0 environment

Symptom

You want to configure a J2EE Cluster for your XI 3.0 system to improve performance in the Java components.
Other terms

Integration Builder, Integration Server, AFW, Adapter Framework, mapping, J2EE server node
Reason and Prerequisites

You want to use additional J2EE server nodes to improve the performance of your J2EE Engine. This can ensure a higher throughput in the mapping step in the XI runtime and a higher throughput in the Adapter Framework.

There are a number of special considerations in this case.
Solution

There are two general approaches to configuring more J2EE server processes:

1. If your host has sufficient resources, you can add additional server nodes to the central instance. This is the preferred configuration, since the costs of configuration and administration are lower in this case.

Note the following:

Each J2EE server node allocates the specified MaxHeapSize and the memory consumption is thus totaled across all servers. Do not set the Heapsize of the individual J2EE server nodes too high in thisregard.

If you use several J2EE server nodes, the number of parallel JAVA mapping connections (JCo RFC provider - destination AI_RUNTIME_) for each server node can be reduced.

If you have already made local changes to the parameter settings in your existing J2EE server node, make sure that these parameter changes are also transferred to the new J2EE server node. The simplest way of doing this is by always making the parameter settings globally (Global Configuration).

If you also use an SLD on your XI J2EE Engine, see also SAP Note 825116 regarding SLD and cluster installation.

With each server node, a new instance of the defined XI communication channels is created. This must be taken into account when you define the channels and program the connection or when you program adapter modules.

Examples:
JMS Adapter - If you use an EOIO sender channel, you must specify the queue ID and the cluster ID of a J2EE server node. For more information, see the SAP Help Portal under http://help.sap.com/saphelp_nw04/helpdata/en/f4

/2d6189f0e27a4894ad517961762db7/frameset.htm

(XI Settings - Parameters: Quality of Service).

File Adapter - see SAP Note 821267
"FAQ: XI 3.0 File Adapter"
- Point 4 : FTP Sender File Processing in Cluster Environment
- Point 7 : File Construction Mode "Add Time Stamp" and "Add Counter" in Cluster Environment

JDBC Adapter - make sure that read entries in the database tables are then either deleted immediately or marked as transferred.

2. You can install an additional dialog instance with the dispatcher and one or more server nodes.

Note the following:

Each J2EE server node allocates the specified MaxHeapSize and the memory consumption is thus totaled across all servers on one host. Therefore, you should not set the Heapsize too high.

When you are patching the XI software in the Offline Deployment step, note that you must stop and restart the J2EE Engines on all instances.

The above points for SLD, AFW, mapping and so on also apply to this installation.



Furthermore, no XI system-specific configurations are necessary, but pay attention to adapter and scenario-specific configurations.

For the sake of completion, we refer you to the following documentation for correct JAVA programming in the J2EE Engine cluster environment:
"Clustering Awareness of Java Applications in a SAP NetWeaver Cluster" under

https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/

docs/library/uuid/e10b4b78-0801-0010-aaba-fb0a5f9c74fc