Embedded Search in different deployment options of a Knowledge Warehouse used by a SAP Solution Manager 7.20 system
There are different options to handle documents in SAP Solution Manager 7.20. The most common option is to run the Knowledge Warehouse in the SAP Solution Manager 7.20 system itself and to store the documents there, using a HANA database. But there are different other options which are discussed here. These options have specific advantages or disadvantages.
It is essential to have a fast access to documents using the "Embedded Search" infrastructure. This infrastructure can work with a HANA database directly, but it also can work with a TREX server, if no HANA database is available. The graphics below shows you the main three variants to run a Knowledge Warehouse with a SAP Solution Manager 7.20 system.
Variant 1 uses a HANA DB for all data in the SAP Solution Manager 7.20. and the KW system runs in the Solution Manager system itself. The Embedded Search infrastructure has access to all data including KW documents. Embedded Search has a deep integration into the HANA database, for that reason we recommend this variant. However, customers are sometimes concerned that storing the KW documents in a HANA database might result in high costs for HANA database memory. These concers are discussed below. Please note that the KW document content itself will not consume HANA memory.
Variant 2 uses a different system for the KW and only the Solution Manager runs on HANA, the KW system runs on a different database. In this option, the HANA database is connected to the embedded search infrastructure. As the KW system runs on a different kind of database, the Embedded Search infrastructure builds an index of the KW documents. Only this index, not the whole document content, is stored in the HANA database and used for fast access to the documents.
Variant 3 is a SAP Solution Manager that runs on a non-HANA database. In this case, a TREX server is needed for the Embedded Search infrastructure. All documents will be read from the database and an index is built by the TREX server to allow a fast access to the KW documents and to all other indexed content.
Questions concerning memory consumption using a HANA database for the Knowledge Warehouse documents
We often get questions that our customers want to use the HANA database for KW documents, but are concerned that this will blow up the memory needed for the HANA database and thus will lead to exploding costs. For that reason, we have put together here the information that shows you that using a Knowledge Warehouse on a HANA database will not store the documents themselves in the HANA database memory. The documents will be indexed instead to allow a fast access. The full document content however will not consume HANA memory directly. Only the index is loaded into the HANA memory.
As this is a very important question, we want to give you some more technical details. The KW documents are stored in so called LOBs ("Large OBjects" that are usually stored as binary data). These LOBs are handled in a special way by HANA: as the HANA database usually cannot use the LOB data in its column store, they are not loaded into the memory. So no memory is wasted for KW documents. Instead, the data is indexed by the Embedded Search infrastructure to have a fast access to this data. But you do not have to add the size of all your KW documents to the HANA memory that is required. This is correctly taken into account by the SAP Quicksizer Tool. This application gives you a rough estimation of the overall size needed for your system. Please refer to the corresponding documentation of the Quicksizer Tool for more details.
As a background information, we want to show you where you find the original source of information that describes that LOBs are not loaded into the HANA memory. This is a generic mechanism in HANA and it is described in SAP note 1872170, there you can look at the FAQ document that is attached to that note. This document has the name FAQ_of_SAP_Note_1872170_v68.pdf. Open this document and you can find the corresponding information in chapter 12.
Questions concerning other non-SAP document management systems
If customers want to use other document management systems instead of KW, they have to use the element type "Document URL" in the process documentation UI. This gives the customers the possibility to store any kind of link. The link itself is stored in the KW, but it points to a document outside the KW. You have the possibility to assign the usual document attributes such as "Document Type", "Status" and "Language" also to these links. However, using such links without storing the document content in the KW implies various disadvantages:
- The documents are not indexed by the Embedded Search. This means that you will not be able to find the documents if you use the search functionality in the process management UI
- Digital signature is not available for such documents, as they are stored outside the SAP system
- "Lifecycle management" does not work the same way. This means that document changes will not produce new versions in the branch, as the node itself (which contains the link) remains unchanged all the time
- History information is not available in the Solution Manager. Changes of the document are not reflected in the "timestamp" and "last changed by" attributes of the respective node in the process documentation UI. Historic versions of the document itself cannot be accessed from the process documentation UI.
If you want to have the full integration of documents into the SAP Solution Manager and its process documentation UI, we recommend to use the Knowledge Warehouse to store your document content. Only in this case you have the full functionality concerning history, lifecycle management and digital signature as well as the full search capabilities, as only then the document content is fully indexed by the Embedded Search.
Question concerning license costs for HANA database in SAP Solution Manager 7.20
Please refer to the following links for any questions concerning license costs for HANA database in context of SAP Solution Manager 7.2
For further details, see https://support.sap.com/usagerights.
Question concerning upgrade to SAP Solution Manager 7.20 and at the same time moving the documents to another content server
Some customers want to change the content server storing their documents and wonder whether they should move the documents to another content server before or after the upgrade.
Moving documents to another content server is a special question that is discussed in SAP notes of the respective product. Those notes handle the way how a new content server is connected to a system and how documents are transferred. Please refer to SAP notes 710711 and 389366 for details.
However, we want to give one recommendation here concerning this question, as there is also one step that handles the documents during the content activation. Some basic settings concerning documents have also changed in SAP Solution Manager 7.20. For that reason, we see a clear advantage if you first do the upgrade and then move the documents to your new content server. Even though both options will be possible, the option that does the upgrade fist, with the old content server connected, causes less effort. The reason is that connecting the new content server before the upgrade will cause the effort to configure this new content server twice - once for the 7.10 storage (area IWBSOLAR) and once for the 7.20 storage (area SMD). This effort can be minimized by proceeding as recommended above.
Question concerning authorizations for documents
please refer to KBA number 2455291
Questions concerning the necessity to use TREX (only in case you have NO HANA database)
please refer to KBA number 2455461
KBA , SV-SMG-IMP-BIM , Business Blueprint / Solution Documentation , How To
About this pageThis is a preview of a SAP Knowledge Base Article. Click more to access the full version on SAP ONE Support launchpad (Login required).
Search for additional results
Visit SAP Support Portal's SAP Notes and KBA Search.