Show TOC

Procedure documentationConnecting To an Embedded Search System Locate this document in the navigation structure

 

You can connect SAP NetWeaver systems on which Embedded Search has already been implemented to SAP NetWeaver Enterprise Search. One example of such a system is SAP Business Suite 7. This allows you to set up a delegated and a federated search.

After you have created the connection, you can use the Move Search function to move the following data and objects from the Embedded Search system to the central SAP NetWeaver Enterprise Search system, also known as the hub system:

  • Modeling data

  • Configuration data (search object connectors, software components, system information, scheduled indexing runs, and extraction jobs)

  • TREX indexes and their content

After completion of the move, the Embedded Search system is no longer available for modeling, connector generation, and connector management. You can now perform these tasks for all connected and moved Embedded Search systems in one central location, the SAP NetWeaver Enterprise Search system.

Prerequisites

  • The Embedded Search system is based on SAP NetWeaver 7.0 with Enhancement Package 1 or higher.

  • You require the SAP_ESH_LOCAL_ADMIN role.

  • If there are already search object connectors in the Embedded Search system, you must create two directories on the TREX servers of the Embedded Search system and hub system. In the Embedded Search system, you require a directory for the export of the index contents and in the hub system, you require a directory for the import of the index content. To create the respective directories, proceed as follows:

    1. In the Embedded Search system, launch transaction SE38 and run report ESH_INT_CREATE_TREX_EXPORT_DIR.

    2. In the hub system, launch transaction SE38 and run report ESH_INT_CREATE_TREX_IMPORT_DIR.

    Recommendation Recommendation

    We recommend that you use the suggested directory names.

    End of the recommendation.

    Note Note

    For the temporary storage of the index content, which is copied to the import directory and the export directory during the move, there must be at least as much disk space on the TREX server for the Embedded Search system and for the hub system as the index folder of the Embedded Search system currently occupies.

    To find out how much disk space is required, check the size of the index folder on the TREX server for the Embedded Search system, which you can find at /usr/sap/<SID>/TRX<INSTANCE_NUMBER>/index by default. This path can be configured, therefore it may have a different name in your system landscape.

    End of the note.

Procedure

Connecting To an Embedded Search System

To create a connection to an Embedded Search system, proceed as follows:

  1. In the hub system, launch the administration cockpit.

  2. Choose   System Settings   Connect Back-End System  .

  3. Select the type SAP BO Search.

  4. Use the Add function to create a new back-end system and enter the required system information.

  5. Choose Maintain Connectivity.

    The hub system creates a connection to the Embedded Search system.

  6. Specify the directory that you created for the export of the TREX data in the Network Folder for Export of the Index Content field.

  7. Choose Move Search.

    All configuration data in the Embedded Search system (search object connectors, software components, system information) and TREX indexes are copied to the hub system. To ensure that you can still make application-specific changes to the model data in the Embedded Search system and that the current model status is visible there, the model data is not deleted there.

    Note Note

    After completion of the move, you no longer use the administration tools in the Embedded Search system for modeling, connector generation, and connector management. You can now perform these tasks for all connected Embedded Search systems in the central hub system.

    In the administration cockpit of all Embedded Seach systems, a message informs you whether or not a connection to a hub system already exists.

    End of the note.

    After completion of the move, you can find all search object connectors and software components that were previously available in the Embedded Search system in the administration cockpit of the hub system.

  8. So that users can always search the latest data during production operations, you schedule regular indexing runs on the hub system.

    More information: Scheduling Indexing

    Note Note

    In addition to the indexing types described there, Embedded Search systems allow indexing in real time for certain types of objects.

    The shortest possible interval for a normal scheduled indexing run is one hour. If you want to include even more up-to-date data in the search, choose the Indexing in Real Time option. This type of indexing updates the data with a delay of only a few seconds.

    More information: Controlling Real-Time Indexing

    End of the note.
  9. On the hub system, assign search authorization to the users of the former Embedded Search system. You can use the standard search user role SAP_ESH_SEARCH or your own roles with restricted authorization.

    More information: Using Roles to Assign Required Permissions to Users and Restricting Authorizations for Searching.

Viewing the Logs

During the move, the current status appears on the user interface.

Alternatively, you can use transaction SLG1 to open the logs for the move. To do this, specify the object ES and the subobject INTEGRATION.

Troubleshooting: Connecting and Continuing the Move

If the connection terminates due to an error or you have manually ended the move, you can use the Reconnect function to restore the connection and use the Move Search function to continue the move.

The system recognizes whether or not steps of the move process (Create Connection, Import Model Data, Move Configuration and Indexes) have already been completed successfully. If the connection was terminated during a step, the system resumes the move starting with this step.

Transferring Model Changes From the Embedded Search System To the Hub System

Although you cannot make any further changes to the models after creating the connection to the hub system in the Embedded Search system, it is possible that automatic updates or imported SAP Notes result in model changes. Therefore, you should check all connected Embedded Search systems regularly for model changes.

More information: Updating Model Data in the Hub System

Information About the Transferred Model Data

The release of the assigned system software component is attached to the IDs of the software components transferred from the Embedded Search system, if one exists. If more than one connected Embedded Search system contains identical model data, they are all assigned to the same software component. However, if there are differences in the model data, the system creates a new software component in the hub system and assigns the model data to this instead. This ensures that the hub system contains the exact model data status that exists in the Embedded Search system. If the release information is identical as well, the software component ID is increased by one.

Deinstalling the TREX Server

The TREX server that was implemented specifically for the Embedded Search system, is no longer required for Embedded Search after completion of the move. The indexing runs centrally on the SAP NetWeaver Enterprise Search system, therefore you can deinstall the TREX server for the Embedded Search system if this is not used for other applications in parallel.