For each component of an integration landscape that creates and persists messages (Integration Server, Adapter Engines and Integration Engines), a unique index can be created that contains all the messages of this component. The SAP NetWeaver search and classification engine TREX is required for this purpose.
To configure a component for index-based message searching, you have to:
· Ensure that the component is based on at least SAP NetWeaver ’04 Support Package Stack 15 or SAP NetWeaver 2004s Support Package Stack 06 (this is also valid for the J2EE Engine, on which your Runtime Workbench is running).
· Establish a communication between the component and the TREX server.
· Set up and maintain the indexing of messages on this component.
To enable an index-based message search in an integration landscape by using TREX, you require a TREX installation of the most current support package stack on a separate server. For more information, see the installation guide of the SAP NetWeaver standalone engine TREX 7.0 on SAP Service Marketplace at https://service.sap.com/instguidesNW2004s ® Installation ® 3 ? Installation ? Standalone Engines.
TREX 7.0 support package stack 09 or above is required if you want to use the fast indexing mode for index-based messages searching. For more information, see Index Administration.
For more information about TREX, see TREX: Technical Operations in Detail
Using TREX for index-based message searching may have an impact on the sizing of your TREX server.
You have to set up the communication with the TREX server separately for ABAP components (Integration Engines) and for Java components (Adapter Engines and Runtime Workbench).
ABAP components communicate with the TREX server by using a separate TREX RFC connection. For more information, see Configuration of the RFC Connection.
1. Call transaction SXMB_ADM.
2. Choose Integration Engine Configuration.
3. Choose Change Specific Configuration Data.
4. Add the parameter TREX_RFC_DESTINATION to the category MONITOR.
5. In the column Current Value, specify the TREX RFC connection and save your entries.
For Adapter Engines and for the Runtime Workbench, the data required for communicating with the TREX server is maintained on the corresponding J2EE Engine.
Use the Visual Administrator of the J2EE Engine to configure the TREX service for each cluster server node as follows:
1. Select the TREX service and enter the TCP/IP connection data for the parameter nameserver.address.
2. Save your changes and restart the TREX service.
Use the Index Administration function of the Runtime Workbench to set up and maintain the indexing of messages on your integration landscape components.