It might be necessary to create a search server relation for communication between an application and TREX. The installation documentation on the application in question will contain information on whether you need a search server relation.
Technical background: The need for a search server relation depends on the version of the TREX ABAP client that is used by the application in question. There are the following versions:
If the application in question uses the SRET package, you must create a search server relation. If the application uses the STREX package, this step is not required.
Creating a search server relation consists of the following:
Creating a Search Server Relation
Field | Entry |
---|---|
Search engine |
DRFUZZY This is the internal name of the TREX search engine. Caution
Make sure that you enter DRFUZZY in uppercase and in the format specified. |
RFC Destination (TCP/IP) |
Name of the RFC destination that you created with the activation type Registration. This entry must tally with the name that you assigned when you created the RFC destination. Tip
TREXDEFAULT_REG |
Description |
Description of the search server relation, for example, Search Server Relation for Retrieval Service. |
You return to the previous dialog box.
The search server relation is then shown as default in the table.
Testing a Search Server Relation
Two entries are listed for the search server relation you created: One for action I (indexing) and one for action S (searching).
The connection with the TREX RFC server and the TREX search engine is established. You can see this in the version information that is shown for the TREX components.
The connection with the TREX RFC server and the TREX search engine is established. You can see this in the version information that is shown for the TREX components.
If the RFC connection still cannot be established, seeChecking an RFC Connection.