Choosing The Right Big Data Platform For SAP Environments
Content
Choosing the right Big Data platform platform for SAP environments
Enterprises using or planning to use SAP applications and databases are beginning to visualize—and, in many cases, to actualize—the dramatic ways in which Big Data analytics can have a positive impact on business operations. Organizations are leveraging new insights, pulled together from a wide range of disparate sources, to more accurately understand customer behavior and engage them in real time to t o maximize protability.
Here’s a real-world example: The University of Kentucky is using a Big Data solution incorporating incorporatin g an in-memory database to signicantly improve retention and graduation
rates, potentially saving the university millions of dollars. The solution extracts data from seven of the university’s systems and analyzes the information in memory. The speed of the in-memory solution has been an eye-opener. “The rst time you query a 45-million-row table, join it with ve other sources, and get a response back in less than
a second, as a data geek, you kind of giggle a bit,” says Stephen Burr, advanced analytics lead at the university. Dell developed the system in collaboration with Intel and built it on the SAP HANA inmemory database appliance. The solution provides the university with the ability to quickly identify at-risk students; tailor courses to match the unique needs of students; improve
retention and graduation rates; and enable students to be more successful through personalized guidance. “The cycle time between question and answer has moved from weeks to seconds with
the help of the Dell SAP HANA solution,” says Vince Kellen, senior vice provost, academic planning, analytics and technologies. Burr adds: “Every 1% increase in student retention we predict leads to an additional million dollars in revenue.”
Why Big Data, why now? Because of results such as this, Big Data is taking its place as one of the transformative technologies in today’s computing paradigm. IDC has characterized Big Data analytics as one of the four pillars of a new era in computing, along with mobile computing, social networking and cloud services. The combination of these technologies will account for 80% of growth in IT spending between now and the end of the decade.1 McKinsey & Co. has characterized Big Data as, “the next frontier for innovation, competition and productivity.”2 What are some of the key factors fa ctors driving this seismic shift? Data proliferation: The volume of data data growth is staggering. staggering. The amount amount of data being
¾
generated every 10 minutes today is equal to the amount of data created from the
beginning of recorded history until the year 2003.3 Approximately 90% of today’s data is of the unstructured type.4 Big Data analytics provides organizations in all industries with the opportunity to harness that data and turn it into useful, actionable and highly valuable information and insight. Increased mobility: Along with the proliferation proliferation of data, data, the availability of of near-
¾
ubiquitous mobility is creating big opportunities not only in the creation of data,
but also in the ability of organizations to respond in real time to information they have gathered and analyzed. For many consumers and workers, mobile devices are becoming the primary tool for connectivity. In addition, many more machineto-machine interactions are being driven by mobility. Organizations that can tap
into these mobile sources and deliver actionable information information are seeing signicant
advantages in productivity productivity,, competitive positioning, responsiveness, agility and protability.
Advanced technologies: Innovative vendors such as Dell, Intel and SAP have worked worked together to ensure that the technologies to enable Big Data analytics are available
¾
and aordable. As just one example, the development of in-memory databases,
particularly the SAP HANA platform, is one of the important driving forces in enabling Big Data analytics. Gathering massive amounts of data from disparate sources and analyzing it in real time requires levels of speed and performance that legacy
databases and memory systems simply can’t deliver. It should be no surprise that the companies identied as “best in class” in deploying data analytics technologies are
twice as likely to use in-memory computing technology as other companies.5 Business opportunities: None of this investment in Big Data would take place, place, of
¾
course, if it weren’t driving signicant opportunities for businesses to innovate the
ways in which they operate and interact with employees, customers, partners and others. Here is a sample of some of the opportunities predicted for Big Data, based on a report from McKinsey6: o A 60% increase in operating margins for retailers. o $300 billion in annual value to the healthcare system in the U.S. U.S. alone.
1
“IDC Predicts 2012 Will Be The Year of Mobile and Cloud Wars Wars as IT Vendors Vie For Leadership While the Industry Redenes Rede nes Itself,”
IDC, Dec. 1, 2011 “Big data: The next frontier for innovation, competition and productivity,” McKinsey McKinsey & Co., May 2011 3 “Big Data or Too Too Much Information?” Smithsonian, May May 7, 2012 4 “Extracting Value from Chaos,” IDC, June 2011 5 “Big Data For Marketing: Targeting Success,” Success,” Aberdeen Group, January 2013 6 Ibid, Footnote #2 2
2
o $600 billion in annual consumer consumer surplus surplus from from using using personal personal location data globally. o €250 billion in potential annual value to Europe’s Europe’s public-sector public-sector administration.
Harnessing the power of in-memory solutions and Hadoop As seen in the University of Kentucky use case, one of the keys to the successful deployment of Big Data analytics in SAP environments is the use of the SAP HANA in-memory data platform. SAP HANA dramatically increases processing speed, empowering businesses with real-time insight and analytics. In addition, the SAP HANA platform supports the Apache Hadoop open source framework, which is important in organizing and collecting the massive amounts of unstructured data that are central to the emerging Big Data paradigm. SAP HANA helps businesses address one of the critical aspects of Big Data, which is fast access to—and real-time analytics of—very large data sets. This allows decision makers to understand their business “at the speed of thought.” When used with Hadoop, SAP HANA can combine, analyze and process all of the data a business has, providing deeper insight into customer behavior and opening up new business opportunities. IT decision makers involved in building solutions incorporating SAP HANA and Hadoop should recognize the dierences between these technologies to understand how, where and
why they complement one another. Hadoop uses commodity servers to handle data in the petabyte and, potentially, the exabyte range, which is much higher than the 100 terabytes t erabytes (TB)-or-less range that SAP HANA and conventional relational database management systems (RDBMS) handle. Current versions of Hadoop are signicantly slower than SAP HANA, or even conventional
RDBMS, taking minutes or hours to provide analytic results. That means that Hadoop, unlike SAP HANA, will not allow you to understand your business at the “speed of thought.” Hadoop does, however, allow you to store and access more voluminous and detailed data at a lower cost. The net result is that by putting SAP HANA and Hadoop together you can maximize the most powerful characteristics of each platform to handle really Big Data really fast. Maximizing the potential of SAP HANA and Hadoop for Big Data analytics also means choosing the platform that will be most eective and cost-ecient for your organization.
In addition to the costs of hardware and software, decision makers must take into account development tools; the operational costs associated with meeting service levels; and how the solution will fulll policies concerning security, high availability, availability, secure backup and recovery.
Advantages of the x86 platform for SAP Big Data environments SAP is typically the touch point for customer interactions and activities; therefore, for any organization using or planning to use SAP, SAP, it often represents the front line in the journey to Big Data success. And if SAP is the pathway to Big Data for your organization, one of the most important considerations is in building your solution on the hardware platform that will deliver the best performance, the most ecient price point and the most readily scalable
architecture for future growth and long-term business opportunities. The x86 platform oers signicant benets for organizations hoping to make the most out of the opportunities aorded them by Big Data in SAP environmen environments. ts. This, of course,
applies to organizations already utilizing the x86 platform for SAP, as well as to organizations
3
SAP. In addition, organizations with aging just starting out with SAP. aging legacy Unix environments environments can easily and cost-eciently transition to t o modern, x86 environmen environments—with ts—with no impact
on applications availability or performance—throug performance—through h solutions such as Dell ZeroIMP ZeroIMPACT ACT Migration Services for SAP. The following are some of the key advantages of an x86 environment: Lower total cost of of ownership: An open open x86-based infrastructure infrastructure and open source operating system such as Linux will result in reduced capital investment in hardware
¾
compared with legacy Unix environments, and will enable signicant additional savings for systems maintenance, software licensing, stang and energy usage. By
utilizing integrated solutions from Dell and Intel, organizations can maximize their savings. An integrated solution comprising high-performance servers and software with a market-leading database can reduce total cost of ownership by 50% or more for SAP Business Suite customers. Better performance: performance: Price is clearly a major major part of of the equation, equation, but so is
¾
performance. Linux systems built on the x86 architecture now outperform Unix when it comes to mission-critical applications.7 That is one of the reasons why analysts project that 65% of applications that were running on proprietary versions of Unix in 2012 will have migrated to x86 by 2017.8 The gap between x86 performance and Unix performance is becoming more pronounced in favor of the x86 environment as Intel continues to innovate. For instance, the Xeon Processor E7 product families
oer breakthroughs in memory, performance and reliability, which will be described in
more detail below. Availability of SAP SAP HANA: The SAP HANA in-memory in-memory database is the linchpin of SAP’s strategy in the Big Data market. The performance gains enabled by in-memory computing are critical in empowering organizations to analyze Big Data in real time and create actionable events using that insight. For any SAP customer pondering deploying the power of real-time, predictive Big Data analytics, SAP HANA will be key: This means that the customer must build the solution on an x86 platform because SAP HANA is only available on the x86 platform using Linux.
¾
Support for Hadoop: Hadoop: Another Big Data capability enabled by SAP HANA is the availability of Hadoop. SAP supports integration with Hadoop in several ways, including the SAP HANA smart data access capability; SAP IQ native MapReduce API; SAP Data Services Hadoop connector; and SAP BusinessObjects BI universes. Hadoop enables organizations to store massive amounts of unstructured data, which, when used in combination with SAP HANA, can deliver real-time analysis and insight.
¾
Advantages of Dell Big Data solutions in SAP environments While it may be clear that x86 should be the platform of choice for Big Data in SAP environments, IT decision makers should also recognize that not all x86-based solutions are created equal. Enterprise IT decision makers will see clear advantages in using the Dell
Integrated x86/ Linux solution builton the latest generation of Intel processors and including support for Intel Distribution for Apache Hadoop. 7
4
“SAP on Linux,” Realtech, Realtech, Oct. 31, 2013 “The Future of Unix: Hazy and Overcast, So Reach for the Umbrella,” Gartner, April 20, 2012
8
Working with integrated Dell and Intel technologies will provide signicant benets in price/performan price/performance, ce,
modular scalability, availability and increased agility—all enabled by technology solutions designed from the groundup to meet the shifting demands of modern data center environments. Most organizations are at the early stages of their Big Data initiatives and will need simple scalability as requirements grow. In addition, the proliferation of
data, both structured and unstructured, is continuing exponentially, creating ongoing challenges—and opportunities—in enabling Big Data analytics to support additional applications. One of the key advantages of the Dell approach is the use of a modular, scalable architecture for SAP HANA. Working closely with SAP, Dell oers a series of optimally congured SAP HANA solutions that consist
of a hardware appliance along with pre-loaded software and a full range of services, including assessment, implementation, management, data modeling and usecase assistance. Dell PowerEdge Servers are the foundational platform for the Dell Solution for SAP HANA and are used consistently across all congurations. Because there are multiple congurations, customers can select a solution
that is right-sized for each data analysis environment, knowing that they won’t have to rip and replace their existing solutions as their needs grow. By using modular scalability as a primary design construct, Dell delivers a solution that grows seamlessly, without requiring a hardware change when expansion is required, and without aecting the performance or availability of the
existing system. In addition to an innovative modular, scalable architecture, Dell’s latest servers—built on the latest generation of Intel processors—deliver immense scalable memory capacity that make them perfectly suited for the capacity and performance demands of next-generation, Big-Data applications (see sidebar, “Servers That Deliver on the Performance of Big Data”). Finally, Dell oers customers a full portfolio of end-
to-end solutions and services in support of SAP HANA applications and Big Data in general through oerings
such as Business Intelligence and Analytics Services. Services complement one another, including assessment, implementation, management, data modeling and use-
Servers that deliver on the performance challenges of Big Data
As enterprises seek to take advantage of the business opportunities aorded by Big
Data, they are realizing that some of their legacy hardware is simply not up to the task. Older servers and storage solutions will often lack the capacity, speed and throughput required for today’s most of demanding applications. While much
the hardware can be repurposed for other applications, it is important to make sure that the solutions targeted for Big Data take advantage of the state-of-the-art technology innovations available today from Dell and Intel. Leading the way in servers for Big Data is the Dell PowerEdge R920, which has an immense scalable memory capacity of up to 96 dual in-line memory modules, providing up to 6TB of RAM capacity. This makes it particularly well suited for a modern data center with Big Data applications built on high-performance, in-memory databases such as SAP HANA. Additionally, the Dell PowerEdge R920 supports up to four 15-core processors to accelerate application performance and data access. The Dell PowerEdge R920 supports up to 24 internal storage drives, including up to 8 PowerEdge Express ash drives. Using ash drives with Dell Fluid Cache data-
caching technology enables centers to accelerate the storage anddata retrieval of data that is centrally maintained on storage area networks. Bolstering the performance of the Dell PowerEdge R920 for Big Data environments are the latest advances in Intel processor technologies. The server supports up to four processors from the Intel Xeon processor E7-4800 v2 product family. These processors deliver a combination of high-performing computing threads; hardware-based advanced reliability features; and large memory, cache and bandwidth capacity, leading to exceptional levels of scalability, application performance and uptime.
case assistance. For SAP HANA, customers can participate in workshops that develop use case and business case justications, and can take advantage of services such as proof of
concept, modernization, business intelligence consulting and support.
How Dell and Intel support Big Data in a modern data center environment Another advantage of the x86 platform—particularly as delivered by Dell and Intel—is the opportunity to build your SAP Big Data solution within an environment that has been designed to address the needs of today’s t oday’s data centers. What are some of the characteristics of a “modern” data center? They include: A hardware infrastructure that supports the need for increased speed, performance and capacity.
¾
An infrastructure that is open and and standards-based standards-based to enable cost eciencies eciencies and
¾
enterprise-wide integration. A simple-to-scale, simple-to-manage architecture that supports growing growing business
¾
needs for exibility and agility. A data center environment environment that that oers the highest highest levels of resiliency, resiliency, reliability
¾
and high availability to maximize the performance of applications, while eliminating downtime. The requirements of applications are constantly changing, changing, as organiza organizations tions embrace
initiatives such as cloud computing, mobility and Big Data. As a result, Dell and Intel are both committed to expanding the capabilities of their solutions to meet changing requirements
for speed, capacity, agility and manageability. For example, the Intel Xeon Processor E7 8800/4800/2800 v2 product families set new standards in a variety of areas that will help to deliver improvements in the performance of Big Data analytics. Using the latest Intel processors, systems are capable of delivering 6 TB of RAM in a 4-socket server. These huge amounts of memory can support local processing capabilities in an in-memory database, such as SAP HANA. In addition to breakthrough capacity, these new solutions deliver performance improvements that double the throughput of previous processors, provide up to 50% more cores/threads and deliver up to 25% more cache. It’s not just the processing speed and power that are critical to the success of Big Data SAP initiatives. Other solutions from Intel and Dell include: Intel Distribution for Apache Hadoop, which is the innite store for Big Data and integration with SAP HANA for real-time analysis;
and Dell Active Infrastructure, which enables organizations to expand their use of Big Data analytics without having to change architectures. Dell Active Infrastructure supports future growth because it has been designed to support additional applications as they are released by SAP.
Learning from successful use cases While a lot of the discussion about Big Data is focused on future potential, many organizations are already successfully deploying Big Data analytics in SAP environments to
6
deliver signicant business benets. The previously mentioned case study on the University of
Kentucky is a prime example. Another compelling use case is provided by Dell SingleScore for Talent Talent Management, which combines Big Data analytics, social media and mobility to provide a wide range of actionable business intelligence that can, among other things, help employers identify, motivate and retain key employees. Employee turnover is a challenge for every business. According to one survey, the cost of replacing an employee making a salary of less than $50,000 amounts to 20% of the employee’s annual salary. The cost of replacing an executive is more than 200% of the executive’s annual salary.9 The Dell SingleScore Solution uses a variety of technologies, including SAP HANA, Dell Kitenga for Big Data analysis on Hadoop, and Dell Boomi integration to deliver actionable insight from information gathered from structured data from business systems and unstructured data from social media. It provides real-time continuous feedback to employees, managers and executives on employee satisfaction, customer satisfaction, worker productivity, training requirements requirements and a variety of other factors.
Conclusion Big Data is here, it’s now and it’s the future. For organizations already taking advantage of the benets of SAP applications and databases, Big Data is the next logical step in transforming
data into real-time, actionable insight. Companies moving forward are already achieving signicant competitive advantag advantage. e.
Putting in place the right foundation for Big Data is critical. Investing in a breakthrough new solution but building it on a dated, legacy hardware platform doesn’t make sense. The benets of building SAP Big Data solutions on an x86 platform are clear: Better performance,
lower total cost of ownership, improved agility, simpler scalability and the availability of the SAP HANA in-memory platform. Dell and Intel Int el have worked closely together, and with SAP, SAP, to provide technology solutions optimized to deliver industry-leading Big Data solutions on the x86 platform. Integrated Int egrated Dell and Intel solutions provide leadership and innovation in price/performance, modular scalability,, simplied migrations, reliability and availability. scalability
If you’re ready to put Big Data to use in your SAP environment, here is how to get started: dell.com/SAP.
“How much does it cost companies to lose employees?” CBS News, Nov. 21, 2012
9
7
