GRC
HR
SCM
CRM
BI


Article

 

SAP NetWeaver BI Accelerator: Giving BI a Boost

by Henry Morris | SAPinsider

July 1, 2007

The information you need to make the best decisions for your business resides within your SAP NetWeaver BI system. But how quickly can you get to it? With increasing demands for up-to-date information, IT teams have had to create workarounds, such as limiting users to set of common queries, to improve response times. Discover how you can enhance your existing BI investments -- with lightning-fast response times -- using BI Accelerator.
 

 

A key reason why companies invest in enterprise application suites is to provide better access to information. Yet getting data out of ERP systems and into the hands of employees is not always a cut-and-dried task. Investments in SAP NetWeaver BI address this issue in several ways:

  • The technology provides packaged InfoCubes, reports, and queries that facilitate end-user access

  • Data transformations — from the SAP ERP source or other source systems to the BI target — enable users to get the information they need, representing substantial value to SAP customer sites

  • Embedded BI provides analytic insights within the context of a business process or task

For these reasons, the penetration of SAP NetWeaver BI continues to increase each year. In fact, SAP's enterprise BI system is the most deployed SAP NetWeaver technology today.

As with any system, in the face of an increasing user base and greater user demands, enterprises are often challenged to keep up. In the process, they have developed some less-than-ideal workarounds to manage the new demands on their systems as they look to:

Maintain performance and response time. To handle increasing numbers of BI users and their information requests, many SAP NetWeaver BI and SAP BW customer sites report the need to build aggregates, collections of data optimized for specific end-user queries that can be foreseen and planned for in advance — weekly reports for revenue by region, for example, or planned versus actual sales. The aggregates are summary data structures in which data from a set of records is precalculated and stored in the data warehouse, improving performance and response times when end users run queries. But loading the aggregates, whether it's done nightly or weekly, represents significant overhead for data and BI administrators. One customer reported that this job takes up 80% of data administration personnel's time — and these aggregates address only a portion of end-user access needs. Maintaining aggregates for predictable queries also limits end-user flexibility to pose ad hoc queries as business situations arise.

Support a wider range of user demands. It's not possible to predict all of the questions that end users will pose. Suppose there is a sudden spike in demand for a specific product in one region, or that a sudden increase in product defects surfaces at the manufacturing plant level. Users will want to explore the root cause of these new trends so they can recommend a course of action to correct the problem. It's not acceptable, from a business perspective, for end users to be bound to an established set of queries, waiting for IT to build the data structures (i.e., the aggregates) to support new queries.

Support high availability and instant access to information. In addition to building aggregates, some BI customer sites have had to install query governors — limits on the number of records an end user can access — to help improve BI flexibility. Others have had to load the data warehouse less frequently (weekly instead of nightly, for example) to avoid the overhead of doing all of the precalculations every night. This is not the ideal approach, of course, since users would prefer to have the freshest and most comprehensive data for their queries.

All of these are clearly compromise strategies, and enterprises are starting to seek alternatives to managing the expanding user base of their BI solutions.

 

Why the Buzz About BI Appliances?
Recent Findings from IDC

Business intelligence appliances — also known as data warehousing (DW) appliances — are gaining acceptance in the marketplace; early surveys show strong interest in, awareness of, and adoption of these appliances.

IDC sought to understand the most important factors behind this acceptance. A recent IDC survey1 posed the following question: "A data warehousing solution usually includes separate server, storage, and software components, sometimes provided by different IT vendors. Would you be interested in evaluating and purchasing a data warehouse appliance from one IT vendor that combines server, storage, and software in a single preconfigured bundle?"

One-third of the respondents either had a BI appliance or were planning to evaluate one within the next 12 months. That's a strong showing and a definitive indicator that BI appliances are gaining market acceptance.

IDC believes that adoption of BI appliances is not just a passing trend. For the foreseeable future, we expect BI appliances to be deployed alongside traditional DW solutions at large organizations and in some cases as new primary DW solutions, especially at midsized organizations that may not have the resources for traditional DW solution acquisition, deployment, and maintenance.

The Trend Toward BI Appliances

An approach that is gaining favor with SAP NetWeaver BI sites is the use of SAP NetWeaver BI Accelerator, a dedicated business intelligence appliance that provides significant improvement in BI query performance while eliminating the need for defensive IT strategies that compromise business flexibility (see sidebar).

A BI appliance is a combination of hardware and software capabilities dedicated to a specific type of work: the reporting and analysis of business data. An appliance, optimized for a specific function, has the potential to do the job faster and more cost-effectively.

BI appliances are making an increased impact amongst a sizable market segment as more end users evaluate, adopt, and deploy these solutions in support of their business's existing analytics needs. These customers are looking for:

  • Lower total cost of ownership

  • Preconfigured solutions with lower maintenance requirements

  • Highly scalable business analytics platforms to address growing data volumes

Accordingly, many SAP NetWeaver BI sites are currently evaluating and deploying SAP NetWeaver BI Accelerator.

SAP NetWeaver BI Accelerator

SAP introduced its SAP NetWeaver BI Accelerator as an appliance that combines software and hardware in one package. BI Accelerator is a highly scalable analytic server that processes queries initiated by SAP NetWeaver BI users — without having to alter existing applications.

The features that make BI Accelerator highly scalable include its use of SAP's TREX search technology in conjunction with the blade server architecture provided by its hardware partners. To create the BI Accelerator appliance, SAP partnered with Intel, which provides the processors, and HP and IBM, which provide their respective server and storage technologies.

SAP NetWeaver BI Accelerator is not a required component of SAP NetWeaver BI. Rather, it is a scalable analytic processor that complements existing investments in SAP NetWeaver BI. Queries continue to go against SAP BW as they did before. But now SAP BW can hand off the execution of those queries to SAP NetWeaver BI Accelerator, where its data compression and in-memory indexing greatly accelerate the response (see Figure 1). BI Accelerator is a supplement to — not a replacement for — SAP NetWeaver BI.

 

Figure 1
SAP NetWeaver BI Accelerator is a "snap-in" appliance

 

Scenarios for Using BI Accelerator

Although SAP NetWeaver BI Accelerator is optional for SAP NetWeaver BI users, it makes sense to consider it in scenarios such as the following:

You need business flexibility — Since SAP NetWeaver BI Accelerator can dramatically improve response times, it provides greater flexibility to the business. Now questions can be asked that would have been cost-prohibitive before. Users can dig deeper into the information and derive a new level of insight.

You need scalability — The appliance can scale up easily by adding server blades or installing an additional appliance as data volumes and the number of end users grows. The improved scalability and performance is additive. No modifications to the InfoCubes or analytic applications running on SAP Business Information Warehouse (SAP BW) are required, enabling more frequent refreshes to the data warehouse.

You need IT agility — By not building aggregates, IT departments save on resources formerly dedicated to query tuning. Equally important, data models can be simplified so that IT can respond much more rapidly to changing business requirements.

IDC has spoken with several SAP customer sites that have implemented this appliance and succeeded in improving the performance, availability, and productivity of both IT and end users — eliminating the need for restrictive IT strategies, such as maintaining aggregates and limiting the number of rows end users can access in a query.2 End users have reported ten-fold and higher improvements in response times when BI Accelerator was added to an existing SAP NetWeaver BI or SAP BW implementation.

In addition, the need to prebuild aggregates within SAP NetWeaver Business Intelligence goes away with the use of BI Accelerator. This frees IT resources to work on value-added, support activities and enables users to pursue ad hoc queries and analysis.

 

End users have reported ten-fold and higher improvements in response times when BI Accelerator was added to an existing SAP BW implementation.

Conclusion

SAP NetWeaver BI Accelerator addresses an important need: extending the value of investments in SAP NetWeaver BI by expanding the types of queries and the number of users that can be supported. With the BI performance boost that BI Accelerator provides, IT can lift restrictions on how many records can be queried and how often the data warehouse must be refreshed with updated data.

With the shortening of decision cycles, users need to detect changes to business conditions quickly in order to respond rapidly and in accordance with business strategy. SAP NetWeaver BI and BI Accelerator are a technology combination that addresses requirements for more end-user flexibility to query and understand the current state of the business.

 

Benefits of SAP NetWeaver BI Accelerator

  • Faster, more flexible query processing and response time, using in-memory data caching for higher performance and permitting queries on all of the data rather than just a subset

  • Faster load times, so more frequent loads are feasible to provide more current data for end users while still ensuring good system availability

  • Attractive packaging as an appliance that is prefigured for analytic processing, providing single vendor accountability


1 IDC and DM Review, Business Analytics Survey (November 2006).

2 See the 2007 IDC report "Business Intelligence Accelerator: Enabling Organizations to Be More Responsive," available at www.sap.com /platform/netweaver/brochures/informationmanagement.epx.


 

Additional Resources

"BIA Saves BI at Kimberly-Clark" by Evan J. Albright (SAP NetWeaver Magazine, Summer 2007, www.NetWeaverMagazine.com)

"Turbocharge query and analytic performance with SAP NetWeaver BI Accelerator" by Storm Archer (SAP NetWeaver/BI and Portals 2007 conference presentation available on take-home CD, www.sapnetweaver2007.com)

"Speed, Scalability, and Flexibility All at Once with New High Performance Analytics" by Andrew Ross (SAP Insider, October-December 2005, www.SAPinsideronline.com)

SAP Business Intelligence (BI) by Norbert Egger, Jean-Marie Fiechter, Sebastian Kramer, Ralf-Patrick Sawicki, and Stephan Weber (SAP PRESS, http://store.sapinsider.wispubs.com)

 

Henry Morris is Senior Vice President for Software and Services at IDC, a worldwide research firm with headquarters in Framingham, Massachusetts. He started the Analytics program at IDC and coined the term
"analytic applications" in 1997. He has written extensively on analytic applications, decision-centric business intelligence, and the ROI of analytics projects. Dr. Morris earned his bachelor's degree with distinction from the University of Michigan and his Ph.D. in philosophy from the University of Pennsylvania. He can be reached at hmorris@idc.com.

 

An email has been sent to:






More from SAPinsider



COMMENTS

Please log in to post a comment.

No comments have been submitted on this article. Be the first to comment!


SAPinsider
FAQ