SAP NetWeaver Business Intelligence Accelerator on Sun

Increase Your Query Performance 10 to 100 times

To increase profitability and control costs, you need to rely on key business intelligence systems like SAP NetWeaver Business Intelligence (BI) to search for new opportunities from within the huge amounts of data your company produces. However, the growing number of BI users, ballooning amount of data, increasingly sophisticated reporting tools, and the need to execute more ad-hoc queries all impact response times.

To help organizations overcome these issues, SAP and Intel co-developed an appliance-like solution to accelerate SAP NetWeaver BI queries. Sun offers the solution for the SAP NetWeaver BI Accelerator on the Sun Blade 6000 family with Intel Xeon server modules, Sun StorageTek 6140 arrays, a shared file system, and Fibre Channel (FC) and networking switches. Built with open hardware and OS, business process management functionalities, and BI, the SAP NetWeaver BI Accelerator integrates easily into existing SAP environments without changing the data model.

SAP NetWeaver BI Accelerator on Sun:

• Pre-tested, optimized, and certified by SAP
• Optimized for performance, scalability, and reliability
• Open standards based solution that integrates easily into existing environments
• Pre-installed, cabled, labeled, and ready to turn on
• Includes SUSE Linux 24/7 service
• Includes SunSpectrum Platinum service plan

How It Works

The accelerator is divided into partitions, one per CPU, to spread the work across the blade system. At load time, the BI server copies data to the accelerator where it is stored in partition files in the accelerators shared file system. At query time, the software caches the information cube data and indexes each partition into the memory of every blade, creating a powerful in-memory database. If new queries involve the cubes that are indexed, SAP sends the request to the SAP NetWeaver BI Accelerator to process. Querying and loading are divided into multiple steps, which run concurrently as separate processes. Each step executes independently, working on a separate piece of the data that is pre-loaded into memory. These multistep, parallel processes enable queries to complete in a fraction of the time it would take to complete on a single system.

Benefits:

• Minimizes or eliminates the cost of aggregates
• On-the-fly aggregation eliminates the need for query tuning
• 10-100 times increased ad hoc query performance
• Users can browse data without IT tuning
• Better, timelier insights and decisions can result in productivity gains.

For more information see:

• Feature Story —Turbocharge Your Queries with SAP BI Appliance on Sun