Netweaver BI Accelerator
manfred | 31 May 2006
Businesses are starting to rely on business intelligence solutions more and more. In fact, SAP has stated that all reporting from their core ERP systems will be moved into the Business Warehouse in future. That means more data and more time spent searching through it. Now, with the co-operation of Intel, Hewlett-Packard and IBM, SAP have developed the Netweaver BI Accelerator. Sounds fancy, but it’s really nothing more than a dedicated piece of hardware with blade servers, external storage and a fast indexing engine.
In traditional BI, the searching and aggregation of data is a lengthy and tedious process. The difference in database design and configuration for OLAP and OLTP applications used to be a big deal. Nowadays, hardware capabilities are such that these design considerations matter less and less. The BI Accelerator appliance available from HP and IBM comprises a large expanse of disk that stores an index if all the data in the data warehouse. The more blades the appliance has at its disposal, the quicker the processing of the index. Running Linux and the SAP search and classification engine TREX, the net effect is a Google for your enterprise. Any search result that can be satisfied directly from the BI Accelerator’s cache is rerouted to it, and all other search and aggregation queries run the traditional route of being processed by the BI backend.
The diagram below is from an IDC white paper:
-
1. Data is loaded from source systems into an SAP InfoCube.
2. An index is built for this InfoCube and stored inside the BI accelerator appliance. These are search engine indexes built using SAP’s TREX search technology. They are stored in a file system (not a database system) using vertical decomposition (a column-based approach as opposed to the row-based approach that requires more read time). This results in highly compressed data sets that further contribute to fast processing speeds.
3. BI accelerator indexes are loaded into memory where the query is processed. In memory, joins and aggregations are done at run time. Loading of indexes into memory happens automatically at first query request, or it can be set for preloading whenever new data is loaded.
4. At run time, query requests are sent to the analytic engine, which reroutes the query to the BI accelerator.
5. Query results are returned to the end-user application.
Note:
Steps 1 through 3 above typically are performed offline, (e.g., during less critical times).
Steps 4 and 5 below are executed at actual query time.
So, the trend in SAP continues. More and more hardware and more and more complex system environments. The hardware vendors have every reason to smile! More information on the BI Accelerator here.
- Comments
- No Comments »
- Categories
- Netweaver
Comments rss
Trackback
Alonso continues his blistering pace and maintains his ability to stay ahead of the pack. Monaco, a very twisty circuit leaviing little overtaking opportunity, became the latest race won by the Spaniard. This time, a safety car phase prevented Räikkönen from launching a sustained attack. The McLaren’s engine didn’t last, giving Alsonso plenty of time to make it to the finish without too much stress.
Recent comments