Petya Petrova - Fotolia

Using SAP Hana as a data warehouse

Enterprises are using SAP Hana for in-memory data marts and SAP Business Warehouse implementations that integrate with other data warehouses

Today, SAP Hana can be used in a wide range of capabilities and use cases, and it is fair to say that SAP is betting its future on the success of its new flagship technology.

Of course, there are a number of other in-memory products available in the market, such as those from Oracle, IBM, Microsoft, Pivotal and Teradata, and analytics appliances as well as integration middleware or cloud services.

The most common SAP Hana use case scenario in the market today is as a high-speed in-memory data mart.

Other use cases include SAP Hana as a database, analytics appliance, integration middleware, application server, development environment and a cloud platform.

In many straightforward scenarios, customers of SAP Hana continue to do the same reporting on the same set of data sources but with significant acceleration, by a factor of 100 to 1,000-plus, depending on the specific scenario.

In most cases, customers are not replacing their existing large data warehouse with SAP Hana because of cost and limitations on large-scale data warehouse implementation, but are complementing it to support near real-time analytics and new business insights. However, other examples exist where some customers have consolidated their heterogeneous smaller data warehouse landscape on SAP Hana, to the extent that sizing allows.

Today, the largest SAP Hana implementations go beyond 10TB, although most are between 1TB and 5TB. SAP has internally tested SAP Hana to scale to 100TB.

Capabilities and differentiators

SAP Hana is a fully fledged database management system, one of the leading in-memory column-store relational database platforms that run massively parallel across multiple nodes in a clustered configuration. Unlike many other products in the marketplace, Hana is not a bolt-on indexing solution to enhance what is essentially a disk-based, OLTP database engine.

Read more on big data appliances

SAP customers have a lot of work to do make the most of their SAP environments before they can move to Hana-based systems

The case for commodity hardware vs. Oracle Exadata architecture

Teradata is the latest technology vendor making a play for businesses looking for easy-to-implement big data apps

SAP Hana was built from the ground up as an in-memory system to address challenges that customers were facing with their growing SAP business application and business intelligence (BI) platforms. However, the limitation is that if it cannot fit into dynamic random access memory (DRAM), SAP Hana cannot run; it cannot have partial data on disk and memory. SAP Hana allows only for complete data in-memory.

Enterprises are using SAP Hana for in-memory data marts and SAP Business Warehouse (BW) implementations that integrate with other data warehouses, including SAP Sybase IQ.

According to SAP, Hana has more than 3,600 customers today. However, Forrester estimates that one-third of these are production customers, most of which are using it as a complementary in-memory data warehouse. The most dominant licensing of choice is the SAP Hana Enterprise Edition. However, customers explicitly interested in running Business Warehouse on SAP Hana can license Hana in two ways:

  • By units of 64GB. This is priced as a flat fee per unit for up to 10 units, and the price then decreases with every additional 10 units a customer buys. Future licensing purchases are accretive and retroactive.
  • By software application value. Customers pay a percentage of their total SAP purchase price and receive an unlimited

Runtime licence for Business Warehouse is a licensing option that is also available, including enterprise resource planning (ERP) on Hana. For enterprises that want to leverage enterprise data warehouses on a cloud platform, SAP Hana is available on Amazon Web Services, although currently most are small to mid-sized deployments that are mainly less than a few terabytes in size.

Molson Coors Brewing is a Hana data warehouse user. It uses Hana as a data warehouse to simplify and accelerate its business intelligence architecture. Molson Coors Brewing did not initially redesign any reporting; it moved BW to the Hana platform to run its reports more quickly.

"Specific to Molson, we didn’t redesign the solutions, although we know that others have," says a BI specialist at Molson Coors Brewing. "The jobs we ran were over-running our overnight processing window. We still use ETL [extract, transform and load] and the Hana migration was a simple technical upgrade, since we don’t directly report on any Hana views. At this point in time, we don’t do real-time reporting...it’s not in the near future for us."

Systems integrators with experience in migration are among the routes firms can take to deploy Hana.

Syntel, for instance, helped an insurance firm with more than five million records in BW upgrade and convert the existing system to Unicode (SAP NW BW 7.31 to SAP NW 7.4) to migrate from Oracle to Hana Database, redesign and optimise the huge database and applications to SAP BW powered by Hana, and to deliver a unified solution for Business Warehouse and planning and consolidation on SAP Hana.

Another example is HCL Technologies, which helped a global consumer goods company that is over 100 years old and whose double-digit growth in recent years required a new organisational structure. Its BI solution was unable to adapt to the new structures or cater to the increasing data sources that had to be integrated.

"The project justification of 'pain points' will always come from business," says a BI project manager at HCL Technologies. "It’s rarely just the speed of reporting."

Advice on deploying Hana

Application development and delivery (AD&D) professionals need to understand the full spectrum of Hana capabilities and use cases and build a long-term vision and roadmap for their company that leverages investments and provides a platform for real-time business transformation. If you start by accelerating your existing reports using Hana as an in-memory data mart, consider, as a next step, using it as an analytics appliance for some advanced analytic technologies, such as predictive, that run natively within Hana.

SAP Hana can be expensive, and organisations need to understand the return on investment before committing to a long-term strategy. AD&D professionals need to understand what licence their company currently holds for SAP Hana (if any) and what licence is required for the first use case to get started.

IBM, Microsoft, Oracle and Teradata also offer in-memory database options that are viable for non-SAP sources. These alternate solutions offer in-memory solutions to deliver high-performance analytics and operational reporting.

This article is an extract of the Forrester report SAP Hana is the answer! What’s the question? (November 2015), by Holger Kisker, George Lawrie and Noel Yuhanna.

Next Steps

Learn more about SAP's IQ column-oriented database

SAP releases BW/4HANA data warehousing software for HANA

Read more on Big data analytics