/Project Management/HANA
Take an introductory tour through the SAP HANA environment from a consultant’s perspective. Find out the details behind SAP HANA 1.0 and SAP HANA 1.0 Support Package 03, how the new technology differs from existing SAP offerings, and potential scenarios for implementing SAP HANA.
Key Concept
SAP HANA 1.0 Support Package 03 allows an SAP NetWeaver BW 7.3 system to be run entirely on a HANA database. This set-up potentially offers substantial performance improvements for both data reporting and loading.
SAP HANA — SAP’s in-memory computing platform — is a new technology that has people talking. Technology changes mean that new software and techniques must be developed, and SAP HANA is the culmination of at least 10 years of technological change. Computers have changed over time in line with Moore’s law about trends in computing. In 2002, CPU speed stopped increasing (3 GHz) and progress continued elsewhere. Main memory and the number of CPUs increased. These trends are continuing and have pushed software developments in alternative directions to make use of these improvements.
SAP HANA brings the potential to improve corporate systems on a massive scale — and on a level similar to the move from mainframe to client-server, or magnetic tape to disk storage. For example, in the long term, SAP HANA could:
- Allow real-time reporting on transactions posted within current SAP ERP Central Component (ECC) technology
- Support processing of transactions (although this introduces many complexities and technical challenges, such as table-locking and resilience). A single SAP HANA system could deliver both business transactions (currently provided by SAP ECC) and analytics (currently delivered by SAP NetWeaver BW).
SAP HANA centers on in-memory computing technology. The term HANA refers to both a technology and product. As this can be confusing, within this article I refer to SAP HANA technology or SAP HANA products. To make things more complex, SAP has now released a second version of the product (SAP HANA 1.0 Support Package 03), which contains functionality that is very different compared to the first version (SAP HANA 1.0).
For more information about the nuts and bolts of SAP HANA, see the sidebar, “An Explanation of SAP HANA’s Technology” at the bottom of this article.
Business Benefits Behind SAP HANA
Over the coming years, SAP HANA technology will substantially change the enterprise system’s landscape. Because the technology is new, it is difficult to predict with certainty how existing products will change and new ones will become available.
However, even early on, it’s clear that SAP HANA will allow a business to do things that were not previously possible, offering the following improvements:
- Massively improved reporting performance
- Real-time extraction on a large scale
- Within a given time frame, handling a much greater volume of data
SAP HANA offers the opportunity to gain business advantages by tackling challenges that were previously impossible. For example, a typical business will ship products and then access key performance indicators (KPIs) some time later, as data must be collected and aggregated, and measures must be calculated within the system. However, SAP HANA allows these KPIs to be calculated and re-calculated in real time, as the shipments occur. This lets management see up-to-the-minute KPIs, and allows for immediate updates as and when needed.
Only thinking of SAP HANA as a way to speed up your existing reporting systems is missing the huge potential of the technology: It will allow you to do things that were previously off limits. However, don’t get carried away. Understanding the business processes and requirements remains critical and continues to require expertise and significant involvement. You must still design the system to effectively address the requirements and deliver value. The greatest challenge is always to ensure the successful adoption of the system and making sure its full benefits are realized. SAP HANA is expensive and requires new hardware and new licenses. What’s more, early implementations usually have to address many issues associated with the early adoption of a technology.
Overview of Currently Available SAP HANA Versions
Let’s briefly look at the two SAP HANA versions on the market:
SAP HANA 1.0
In June 2011, SAP HANA 1.0 went into general availability. This major milestone means that customers can now use the product in live production environments. SAP HANA 1.0 is an analytical appliance that is designed to provide super-fast reporting. The product includes an in-memory database, analytics engine, reporting interface, and tools to get data in and out of the database.
SAP BusinessObjects BI 4.0 is recommended as the tool to provide the reporting interface. SAP BusinessObjects Data Services gathers the data from multiple sources. Possible sources of data are SAP and non-SAP systems, which include major relational database management systems, Web services, and flat files.
SAP HANA 1.0 requires specific certified hardware. SAP is keen on early implementations being successful, so it has set stringent criteria for the hardware that can be used. Essentially, SAP is trying to ensure that SAP HANA is seen in its best light in these early days.
SAP HANA 1.0 Support Package 03
Previously known as HANA 1.2 and HANA 1.5, the final name given by SAP to this version is SAP HANA 1.0 Support Package 03.
In November 2011, this version went into Ramp-Up, which is a limited release of the software to approved customers. SAP does not permit these customers to use the version in a live production setting until the later general availability release occurs.
SAP HANA 1.0 Support Package 03 is a high-performance alternative to the normal database on which SAP NetWeaver BW runs. Currently, companies can only use SAP NetWeaver BW 7.3 with SAP HANA 1.0 Support Package 03. It replaces the normal Microsoft, IBM, or Oracle database.
As it uses both in-memory and column storage, this version of SAP HANA offers a performance advantage to the traditional choices. It accelerates data loading, reporting, and planning applications.
Note
For more information about the deployment strategy for SAP HANA, click here to read a roadmap from SAP:
https://tinyurl.com/7uym587 How SAP HANA Affects Existing BI Technologies
Current SAP HANA products are not intended as replacements for SAP NetWeaver BW. SAP will either adapt existing products to take advantage of SAP HANA as a database or re-write entire applications to use SAP HANA. For example, entire SAP ECC modules could be re-developed to use SAP HANA, or an existing product could be implemented on an SAP HANA database.
Before your company begins considering SAP HANA, be sure it has the following:
- Complex or advanced requirements (e.g., real-time needs)
- Large databases
Current BI technologies remain critical parts of an SAP landscape. Current SAP HANA products do not affect the viability or implementations of any of the products within the SAP BusinessObjects BI suite. The impact on the existing core BI products follows below:
SAP NetWeaver BW
SAP NetWeaver BW has no current end-of-life dates. SAP continues to invest in the system and has made a substantial investment with version 7.3. BEx continues to provide a basic reporting tool that is included within the SAP NetWeaver BW license.
In the future, I expect that many SAP NetWeaver BW systems will run using SAP HANA databases. However, many current versions (i.e., 3.0, 3.5 and 7.0) are not suitable for migrating to HANA, so a large proportion will continue to use non- SAP HANA database technology indefinitely. My personal prediction is that in five years time, 25 percent of SAP NetWeaver BW installations will run using SAP HANA.
SAP BusinessObjects
The first truly new and integrated SAP BusinessObjects tools are now available as part of SAP BusinessObjects BI 4.0. These are a critical part of the SAP HANA 1.0 product. SAP BusinessObjects tools will benefit from SAP HANA by taking advantage of the super-fast performance offered. SAP HANA is not a competing technology to SAP BusinessObjects, and they sit alongside each other in SAP’s long-term strategy.
SAP NetWeaver Business Warehouse Accelerator (BWA)
The BWA technology is similar to SAP HANA. Both make extensive use of memory for database storage and extensive CPUs. However, the products have differences: BWA is an extra system that you can quickly and easily install and connect to SAP NetWeaver BW. SAP HANA is a platform containing a database that you can run applications on, such as SAP NetWeaver BW. While they use the same technology, SAP HANA can be the underlying technology that a SAP NetWeaver BW system runs on, whereas a BWA is a separate system that augments an existing SAP NetWeaver BW system, improving its performance.
BWA will continue in its current form and will be important for SAP NetWeaver BW versions 3.5 and 7.0. If you use SAP NetWeaver BW 7.3, you need to choose between acquiring BWA and moving to HANA. Considerations such as your existing hardware and timescales to deliver a complex project may influence your decision.
When choosing between SAP HANA and SAP NetWeaver BW, consider the following:
- The expected life of existing SAP NetWeaver BW hardware, as SAP HANA requires new hardware.
- The version of SAP NetWeaver BW that you currently run (SAP HANA requires SAP NetWeaver BW 7.3 Support Package 5).
- Your organization’s willingness to accept the risk of running early-version software
- Your need for improvements to SAP NetWeaver BW reporting, or the entire extraction, transformation, and loading tool. BWA will only improve reporting performance, whereas SAP HANA will do this and accelerate data-loading.
- The size of your SAP BW database. SAP HANA is not yet suitable for huge databases.
Scenarios for Implementing SAP HANA
I have talked about SAP HANA technology, what it can do for a business, and what is contained in the currently available products. Let us now consider real-world scenarios in which organizations can deploy these products.
SAP HANA 1.0 Implementations
Companies need to identify substantial business benefits to justify deployment of this product. A significant financial commitment from an organization is necessary, as you need to purchase new hardware, acquire new licenses, staff a project team, and resolve teething problems. The product is ideally suited to a business scenario that must process large amounts of data and provide very quick reporting. For example, it suits a retail operation wanting to report on sales in real-time during a trading day, or a consumer goods company wishing to track projected inventory levels using the orders placed on that day.
In addition to learning about SAP HANA, the project and support teams need knowledge of the following tools: SAP BusinessObjects BI 4.0, plus either SAP BusinessObjects Data Services, Sybase Replication Server, or SAP LT Replication Server.
SAP HANA 1.0 Support Package 03 Implementations
SAP HANA 1.0 Support Package 03 is likely to be implemented as a database to run SAP NetWeaver BW 7.3, replacing an existing Oracle, Microsoft, or IBM database. Incidentally, when you deploy SAP NetWeaver BW on SAP HANA, it’s no longer necessary to purchase a license from one of the previously mentioned vendors, which is a useful savings. You will need to store your entire SAP BW database in SAP HANA, which at this stage may make the new technology expensive for customers with large SAP NetWeaver BW databases.
Implementation is a project for a Basis team. An existing SAP NetWeaver BW 7.3 system can migrate over to use the SAP HANA database, or you can create a new empty SAP NetWeaver BW 7.3 system. Migrating an SAP NetWeaver BW system is similar to moving hardware: The database needs to be exported from the existing system, and then imported into the new system.
Ongoing skills are not hugely different to those needed in a normal SAP NetWeaver BW landscape. The Basis team needs to know and understand SAP HANA. However, the required SAP NetWeaver BW skills remain the same.
Consider if a new SAP NetWeaver BW on SAP HANA system will replace or augment an existing SAP NetWeaver BW landscape. You may want to consider a separate environment for high-performance applications that need SAP HANA. Existing, standard reporting could remain on existing lower cost SAP NetWeaver BW hardware. This approach is particularly true before SAP HANA 1.0 Support Package 03 passes the general availability milestone, as it is not yet suitable for live production use.
Comparing SAP HANA and BWA Installations
BWA and SAP HANA share many of the same technologies. They make extensive use of in-memory processing and speed up SAP NetWeaver BW reporting. However, the way they are implemented is very different. As such, it is not possible to move directly from a BWA environment to one powered by SAP HANA.
Installation of BWA is quick, with little configuration needed. BWA runs on hardware called blades, which are re-configured and easily installed into the data center. Once the hardware is installed, it is connected to the SAP NetWeaver BW system, and you nominate which InfoCubes to be accelerated. BWA does not affect the SAP NetWeaver BW system, but instead offers optional add-ons that sit alongside the main system.
SAP HANA is different; it changes the database that sits underneath SAP NetWeaver BW. As such, you cannot implement HANA without fundamentally affecting your SAP NetWeaver BW system. Changes such as this require substantial planning, testing, and coordination.
Final Thoughts
At the moment, you are seeing only the very first of the products that use SAP HANA and its first customer implementations. In the coming years, expect the number of available products and the amount of customers using SAP HANA to increase substantially. SAP HANA technology will change the SAP landscape and potentially affect the entire SAP product portfolio. Companies will be able to tackle business problems that have previously been unaddressed. For example, while negotiating with a customer, SAP HANA allows a business to accurately project the profitability and margin of the sale, while taking into account and using information on stock levels, raw material replenishment costs, and other potential orders.
Key questions that I still have about the currently available SAP HANA versions include:
- Will the promised advantages of performance materialize?
- What will be the costs for SAP HANA running on SAP NetWeaver BW compared to a traditional SAP NetWeaver BW installation?
- What level of problems can early adopters of SAP HANA expect?
An Explanation of SAP HANA’s Technology
SAP HANA uses a different type of database compared to that which SAP normally uses. This fundamentally changes the way data is stored. Companies are used to storing data — a single piece of data is considered an ordered list of elements and is also known as a tuple — in a row, with a series of columns or attributes storing the values. In contrast, SAP HANA uses a column store database, in which a single column stores the same attributes for all tuples, vastly improving the database’s efficiency of reading the data and reducing the storage requirements.
Reading aggregated data from a row store database is computationally expensive. To find out the total value for orders created in a given period of time, you must examine every order to establish if it falls within the time period and its value retained. This requires reading and accessing every single row within the table, which could be several million.
Retrieving the same information from a column store database is much simpler. Only the columns that store the data you are interested in must be read.
SAP HANA also makes extensive use of memory. It keeps its main database within random access memory (RAM), which is far faster than traditional disk storage (i.e., 100 to 1000 times faster). Also, because the data is stored in a column database, fewer indexes and aggregated tables are needed, which results in a reduction in storage capacity requirements (i.e., 10-20 times) while a copy is retained on disk in case of a power failure. The combination of these two changes means that a modest amount of memory within a HANA system can replace a much larger database and offer a huge increase in performance.
SAP will eventually develop brand new applications designed to make maximum use of the very different technologies within SAP HANA, as well as adapt existing applications to make use of SAP HANA technology. This is possible because SAP products are developed to work independently from a database. For example, a company can operate SAP NetWeaver BW on Oracle, MS SQL, or IBM DB2 databases, and in each case, the SAP NetWeaver BW product works in exactly the same way.
SAP HANA technology has the potential to change the SAP landscape. For example, SAP products are currently designed to store multiple versions of similar data. For example, the system of SAP General Ledger Accounting uses many different sets of tables that store similar, if not the same, data. Furthermore, this data is then replicated through to a data warehouse and stored again. Keeping this data synchronized and matching is a significant development and maintenance effort.
SAP HANA technology allows the number of tables to be massively reduced. As reading from the database is so quick, no alternative views and aggregated sets of data need to be maintained. This reduces the size and complexity of a system. Furthermore, reporting on the data held within the application tables can be done within a single environment. It is no longer necessary to replicate the data into a separate system optimized for reporting.
Current SAP HANA products focus on improving access to data for reporting purposes. The technology is not yet mature enough to run a transactional system so, at this point in time, HANA allows you to do what you currently do, only faster.
These advances mean new things are possible. HANA will allow huge volumes of data to be processed rapidly, meaning that systems can undertake more complex processing activities in shorter times. SAP indicates that future versions of HANA products will offer real-time reporting on huge amounts of data, super-fast analytics, a reduction in the size of databases, and a massive increase in the speed of extraction, transformation, and loading functions. This means you can expect to see further improvements in performance for both transactional and reporting systems.
George Campbell-Kelly
George Campbell-Kelly is a certified senior BI consultant at Bluefin, the United Kingdom’s largest dedicated SAP consultancy. Since first working with SAP BW in 2000, he has led the delivery of BW systems that operate with SAP R/3, SAP CRM, and Oracle JDE.
You may contact the author at george.campbell-kelly@bluefinsolutions.com.
If you have comments about this article or publication, or would like to submit an article idea, please contact the editor.
Premium
