Search and Classification TREX has the following new functions in the 7.1 version:
For more information about TREX, see Search and Classification TREX.
Support for 64-bit operating systems to improve performance
TREX 7.1 supports the following 64-bit operating systems:
For more information, see SAP Note 965097, TREX 7.1 platforms, and the Platform Availability Matrix (PAM) on the SAP Service Marketplace at service.sap.com/pam.
SAP determined to optimize the performance of the search and indexing functions of SAP NetWeaver that are provided by the TREX engine. This aim is best achieved by means of the optimum utilization of the CPU architecture. SAP has to focus on optimization measures for a small number of platforms. After carefully weighing up the pros and cons for customers, partners, and SAP; it was decided to concentrate on the Windows and Linux 64-bit operating systems.
The consolidation of platforms is the best way for SAP to guarantee development and support for a TREX version with optimum performance.
The TREX engine is normally executed in client/server mode, whereby a TREX Java client or ABAP client calls the actual TREX server. The TREX server is independent of databases or any other software other than the operating system and relevant communication servers (Web server and Gateway). The new platform strategy therefore has no restrictive effect on other SAP solutions.
Earlier TREX releases (TREX 7.0 and below) continue to be supported on their platforms until the relevant end-of-maintenance date as specified on the SAP Service Marketplace: service.sap.com/pam. This means that TREX 7.0 is supported on all of the original supported platforms until 2014.
Our hardware partners have been informed about the new platform strategy for the TREX engine and will support you in determining the optimum utilization of TREX.
Large TREX indexes with more than 2 GB working memories
For performance reasons, TREX indexes are loaded to working memories. By enabling TREX 7.1 to run on the Windows and Linux 64-bit systems, the restriction limiting working memories to a maximum of 2 GB has been removed. The index size is now only limited by the total amount of available working memory. You can now create indexes that require more than 2 GB of working memory, therefore you can now index and search more comprehensive and more complex data and document sets.
Delta index in working memory for join indexes
When TREX updates an index, it rewrites the majority of the index files. If the indexes are large this process can take a long time and generate a high system load. To speed up the update process, TREX enables the activation of a delta index that can be retained in the working memory like the main index. The delta index is a separate index that TREX generates in addition to the main index. If the delta index is activated, changes flow into the delta index. Fewer files are affected by updates because the delta index is smaller than the main index. The delta index can therefore be updated more quickly. The delta index thus speeds up the updating of the main index. The system integrates the delta index into the main index at regular intervals, thereby keeping the main index up to date. The integration of the delta index into the main index takes place more quickly and without search downtime because both the delta index and the main index are kept in the working memory.
The delta index functions are now also available for join indexes. A join index is an index that consists of several table indexes and metadata. It allows you to search for structured data that is distributed over several table indexes. Join indexes are particularly important when searching for structured data such as business objects and when using TREX as SAP NetWeaver BI Accelerator.
For more information about the SAP NetWeaver BI Accelerator, visit the SAP Developers Network (SDN) at www.sdn.sap.com/irj/sdn/bia.