G00224138

What Does Hana Mean For SAP BW Customers?

Published: 27 December 2012 Analyst(s): Roxane Edjlali, Kurt Schlegel

The availability of Hana, SAP's in-memory database, modifies the data warehousing options for SAP ERP, SAP BW customers or others considering SAP solutions. This research provides guidance for CIOs and IT leaders on how these changes can affect an organization's data warehousing strategy.

Key Challenges

Over time, in-memory DBMS will be used as part of the overall data warehouse architecture. Today, most SAP Business Information Warehouse (BW) powered by Hana customers are putting up to a terabyte of data in memory, but that will increase over time. Big SAP BW shops will likely find value in replacing a traditional relational database management system (RDBMS) with Hana in-memory database, as it reduces administration costs (no aggregates, simplified data modeling) and improves performance for queries, data loads and planning. While Hana is a generic in-memory DBMS, Gartner believes SAP BW powered by Hana remains a data warehouse architecture primarily for SAP customers and SAP data.

Recommendations

Running a proof of concept (POC) is the best way to assess the cost and benefits of implementing Hana. SAP BW Accelerator (BWA) shops purchasing SAP BW powered by Hana should request license credit for past purchases of the BWA. Depending on the value of the Hana purchase, the client will receive a rebate of up to 100% of the BWA purchase price of the software. Heterogeneous shops that don't exclusively run SAP BW should consider Hana in-memory database to accelerate SAP BW or as part of their data warehouse architecture.

As data warehouses evolve toward the logical data warehouse, organizations will need to decide how to best leverage in-memory DBMS as part of their overall data warehouse architecture.

Analysis
Historical Background
Historically, SAP BW has been developed to be the data warehouse of choice for SAP ERP customers. SAP BW has been widely adopted by SAP ERP customers with around 13,000 customers using SAP BW out of the total 50,000 SAP ERP customers. Interactions over the past 18 months with clients indicate that clients valued SAP BW integration with SAP ERP and its packaged content, but clients indicated that SAP BW implementations suffered from:

Performance challenges: Both from a query response time perspective and a data-loading perspective, clients have reported that low latency data-loading requirements were hard to meet. Complex deployments and administration: Most SAP BW implementations are resourceintensive and business users need a lot of help from IT to access the data they need. Complexity in integrating non-SAP data: Customers trying to integrate non-SAP data into SAP BW struggled to integrate it and had to address the challenges mentioned above for an even greater set of data.

As a result, over the same period, Gartner observed through client interactions the following approaches to data warehousing within the SAP ERP installed base:

For customers who had more than 80% of their data feeding their data warehouse coming from SAP applications, SAP BW appeared to be the preferred choice. For customers with less than 80% but more than 20% of their data warehouse coming from SAP applications, many customers chose to implement two data warehouses. SAP BW for SAP data and another data warehouse for the non-SAP data. As the need to bridge the two appeared, some organizations opted to use SAP BW as a staging area and loaded the data from SAP BW into their custom-built data warehouse. In this scenario, SAP BW continued to also be used as the data warehouse for its initial audience. Others opted to integrate non-SAP data into SAP BW. Finally, the remaining SAP ERP customers opted to build a custom data warehouse and to integrate SAP ERP data into the data warehouse using connectors offered by data integration vendors. Some have also been using packaged data mart offerings to extract data from specific SAP ERP modules.

Page 2 of 7

Gartner, Inc. | G00224138

Impact of Hana and Sybase IQ on Data Warehousing Options for SAP Customers
Since 2010, a number of changes in the SAP product portfolio and road map have transformed the SAP landscape, specifically when it comes to data warehouse solutions and DBMS offerings:

The acquisition of Sybase has added Sybase IQ as a valid data warehouse DBMS technology option for SAP customers. SAP Hana DBMS was first released in June 2011. SAP BW powered by Hana was announced generally available in April 2012 and was further enhanced by SAP BW 7.30 SP8, delivered in October 2012. Immediately after the SAP conference, Sapphire, in May 2012, Gartner surveyed approximately 80 SAP BW customers. While this is a small subset of the estimated 13,000 SAP BW customer base, the responses confirm Gartner customer inquiry interactions (see Figure 1).

Figure 1. Question: When do you plan to move to SAP BW powered by Hana? (N=76)

No Plans to Deploy in Production Before 2015

Will Deploy in Production in 2014

Will Deploy in Production in 2013

Will Deploy in Production in 2012

Already in Production 0 10 20 30 40 50

Percentage of Respondents
Source: Gartner (December 2012)

The chart in Figure 1 is deceptive because the "no plans to deploy" option has the most responses, but when the other options are combined, 46% of respondents intend to deploy SAP BW powered by Hana in the next coming two years. This confirms the strong interest and intention in adopting SAP BW powered by Hana in the short to medium term. From our interaction in inquiries, most customers choose a staged approach to mitigate the risks when adopting a new product but also to optimize the cost-benefit ratio by focusing first on the

Gartner, Inc. | G00224138

Page 3 of 7

subject areas where performance issues have been a major challenge to the business. In addition, some customers may combine the phased approach and the side-by-side deployment that allows both the old and the new systems to run in parallel. Many customers consider that the license cost is prohibitive to the broad deployment or that the performance is acceptable either because volumes are small or response times are not an issue (see Figure 2).
Figure 2. Question: How are you planning to migrate from SAP BW on a general purpose DBMS to SAP BW powered by Hana? (N=38)
Big Bang - Move all subject areas at the same time 9%

Phased - Move one subject area at a time 25% Not applicable 52%

Parallel - Run a dual environment 14%

Source: Gartner (December 2012)

If, like many SAP BW customers, you are considering adopting SAP BW powered by Hana, you need to be aware of the following findings Gartner has collected through customer interactions:

SAP BW powered by Hana does perform significantly better than SAP BW running on a standard DBMS (Oracle, SQL Server or DB2). The acceleration in query times can be very impressive for some queries a hundred and sometimes a thousand times faster. However, this performance gain is not consistent across queries. SAP BW engines can only push a subset of the calculations to Hana, hence the difference across queries. SAP BW SP8, released in October 2012, and SAP Hana, SPS5 released in November 2012, are intended to resolve some of these issues and, over time, SAP intends to push more of the SAP BW calculations to Hana. Compression ratios can vary. Compression ratios announced by SAP (4x compression) vary across the deployments and object types. For example, data-staging objects can get smaller compression ratios than InfoCubes.

Page 4 of 7

Gartner, Inc. | G00224138

All DBMS technologies have varied compression ratios. However, because SAP BW powered by Hana is priced by units of 64 GB of memory, the compression factor has a direct impact on the license price. Running a sizing exercise as part of the POC is the best way to know what compression ratio to expect. Another factor affecting the memory sizing is the calculation overhead that needs to be reserved. SAP recommends that only half of the memory is used to store data, and the other half is reserved for the calculation overhead. One frequent question is if an organization runs SAP Business Intelligence (BI; such as BusinessObjects) against BW on Hana, will performance improve? The answer is that it depends. If you have a lot of calculation logic in the universe as opposed to accessing a Business Explorer (BEx) query without needing a universe, running it on Hana may not help because these calculations cannot yet be pushed down into Hana. SAP has said that pushing universe calculations into Hana is on the future road map. This is one reason why SAP is no longer recommending accessing BW via universes. SAP is recommending accessing via BEx queries directly, whether Hana is used or not. SAP has not been discounting SAP BW powered by Hana. Based on publicly available pricing information provided by the U.S. General Services Administration, the SAP BW powered by Hana entry price is $60,000/unit of 64 GB of memory (see www.gsaadvantage.gov/ref_text/ GS35F0406V/0KPEGT.2HCE36_GS-35F-0406V_SAP70.PDF). SAP and SAP partners' service integrators can help with the migration process from SAP BW to SAP BW powered by Hana. If the customer has purchased its ERP DBMS licenses from SAP, then SAP BW powered by Hana is subject to the SAP Application Value (SAV) license cost (see "Choosing the Appropriate Database Management System Platform for SAP") as Hana is then considered an SAP application. To motivate SAP BWA customers to move over to SAP BW powered by Hana, SAP offers a license credit up to 100% of the SAP BWA license spending.

Conclusion
Gartner client interactions indicate that SAP BW powered by Hana is generating a lot of interest within the SAP BW installed base. Many clients have reported important performance gains both in terms of loading and query processing. SAP states that SAP BW powered by Hana has over 180 customers and over 30 live customers. However, in order for organizations to make the most of Hana, IT leaders should consider the following points:

Decide if SAP Hana can support your complete data warehouse architecture:

Evaluate cost. Managing all of the data of the enterprise data warehouse (EDW) in memory may not be appropriate as it may be too costly. Evaluate SAP Hana high-availability and disaster recovery capabilities and decide if they meet your needs.

Gartner, Inc. | G00224138

Page 5 of 7

Consider your mixed workload management requirements and evaluate if SAP Hana meets your needs. Check references with SAP. To date, most SAP BW powered by Hana or SAP Hana DBMS deployments remain small. However, SAP has been actively working on addressing these requirements. SAP Hana clustering capabilities have been available as of November 2011. Further improvements for addressing high availability and disaster recovery are addressed with SAP BW SP8, released in October 2012, and SAP Hana SPS5, released in November 2012.

Run a POC to justify the investment required for the volume of data necessary to be held in memory. This is true when using SAP Hana as a data warehouse platform or when using SAP BW powered by Hana. Organizations must identify the business value of having data stored inmemory. SAP has been working on offering alternative options to managing data in-memory. With SAP BW SP8 and Hana SPS5, SAP has added to SAP BW support for "not active data" and SAP BW Near-Line Storage with Sybase IQ. Understand how new technologies are changing the data warehouse architecture. Generally speaking, new technologies such as in-memory DBMS, managing volume, velocity and variety are affecting the overall data warehousing landscape. Organizations need to understand these changes and evolve their information infrastructure in order to support these new requirements (see "Does the 21st-Century "Big Data" Warehouse Mean the End of the Enterprise Data Warehouse?").

Recommended Reading
Some documents may not be available as part of your current Gartner subscription. "Who's Who in In-Memory DBMSs" "What CIOs Need to Know About In-Memory Database Management Systems" "SAP BW Customer Survey: Plans to Use NetWeaver BW for Their Data Warehouse Architecture" "Does the 21st-Century 'Big Data' Warehouse Mean the End of the Enterprise Data Warehouse?" "Hype Cycle for In-Memory Computing Technology, 2012"

Page 6 of 7

Gartner, Inc. | G00224138

GARTNER HEADQUARTERS Corporate Headquarters 56 Top Gallant Road Stamford, CT 06902-7700 USA +1 203 964 0096 Regional Headquarters AUSTRALIA BRAZIL JAPAN UNITED KINGDOM

For a complete list of worldwide locations, visit http://www.gartner.com/technology/about.jsp

2012 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. or its affiliates. This publication may not be reproduced or distributed in any form without Gartners prior written permission. If you are authorized to access this publication, your use of it is subject to the Usage Guidelines for Gartner Services posted on gartner.com. The information contained in this publication has been obtained from sources believed to be reliable. Gartner disclaims all warranties as to the accuracy, completeness or adequacy of such information and shall have no liability for errors, omissions or inadequacies in such information. This publication consists of the opinions of Gartners research organization and should not be construed as statements of fact. The opinions expressed herein are subject to change without notice. Although Gartner research may include a discussion of related legal issues, Gartner does not provide legal advice or services and its research should not be construed or used as such. Gartner is a public company, and its shareholders may include firms and funds that have financial interests in entities covered in Gartner research. Gartners Board of Directors may include senior managers of these firms or funds. Gartner research is produced independently by its research organization without input or influence from these firms, funds or their managers. For further information on the independence and integrity of Gartner research, see Guiding Principles on Independence and Objectivity.

Gartner, Inc. | G00224138

Page 7 of 7