Index Management Service
The index management service makes documents in CM repositories accessible for search and classification. This is done in a generic way by providing interfaces to index services that implement these functions.
- Search and Classification (TREX) has been installed.
- A service user that has at least read access to all repositories that are to be indexed exists. This user must be specified in the Crawling Service User parameter of the index management service (see below).
- The crawler service is activated so that document collections can be indexed.
- A taxonomy repository is set up so that the TREX classification service can be used.
- The application property service is registered for all repositories that contain documents to be classified.
- In the TREX configuration, you have specified either a TREX HTTP server (for example, an MS IIS with TREX HTTP server extension) or a TREX name server that dispatches requests to (multiple) TREX HTTP servers. The trexmemory query cache must be configured (see Caches and Configuration of the TREX Service in SAP NetWeaver AS ).
The following index services are contained in the standard delivery.
- The TREX search service (trexSearch)
- The TREX classification service (trexClassification)
- The combined TREX search and classification service (trexSearchAndClassification)
- The PCD search service (portalRolesSearch)
The following parameters define the index management service:
Parameters of the Index Management Service
| Parameter | Required | Description |
|---|---|---|
|
Max Transfer Size |
Yes |
The content of documents whose file size does not exceed the specified value is directly forwarded to be indexed by the Search and Classification (TREX) components. For performance reasons, for documents that exceed the specified value only an address is forwarded to TREX. TREX then fetches the content of the document for indexing. Presetting: 10000 bytes. |
|
Index Services |
Yes |
Specifies the index services called by the index management service. Presetting: trexSearch, trexClassification, trexSearchAndClassification |
|
Repository Manager for Which Search Results Are Wrapped |
No |
Specifies the repository manager for which wrappings are generated for search result objects when searching takes place. The wrapping wraps all required properties to the returned search result objects. This increases performance for search queries. You may only specify repository managers of the type CM repository manager or file system repository manager. If you want to use wrapping for repository managers you developed yourself, you must carry out tests first to make sure that searching takes place without errors if you enter your own repository managers here.
If you have developed your own repository managers and specified them here, remove them again if problems occur during searching. |
|
Supported Languages |
No |
Specifies the languages supported by Search and Classification (TREX) in ISO format. Default setting: EN,DE,ES,FR,DA,NL,FI,IT,JA,KO,NB, NN,PT,SV,CS,EL,HU,PL,RO,RU,TR, ZH,ZF,AR,HE,SK,SR,TH,BG,SL,HR, MS,UK,ET,LT,LV,AF,ID,IS,CA,NO,SH You can use the langPropID custom property of an index to assign a document to a particular language index manuallly. For more information, see Creating an Index . |
|
Crawling Service User |
Yes |
User that is used by the index management service when it accesses documents. Default: index_service Caution: Do not change these entries. This user is defined in user management for the Application Server Java and in KM as a service user with read permissions for all resources (see Defining Permissions ). |
|
Enable Dynamic Taxonomies |
No |
This parameter is reserved future use. Do not change the existing setting (false). |
The index management service is preconfigured in the standard delivery. Normally, you do not need to change its configuration.
To check or modify the configuration, choose Content Management → Global Services → Index Management Service.