HANA, SAP’s new high-speed analytics device, and in-memory computing continue to dominate much of the current conversation in the SAP user community, as it did during the recent SAP Sapphire Now conference in Orlando, Fla. At the same time, users continue to ask questions about what SAP HANA in-memory means for them, how it affects their SAP landscapes and what they should be doing about it.
SearchSAP.com spoke to Dan Kearnan, director of SAP business intelligence about HANA and in-memory computing to get answers to some of these questions and learn how the technology affects applications like Business Warehouse (BW) and Business Warehouse Accelerator (BWA).
We’ve heard about SAP BW riding on top of SAP HANA. Can you explain what the roadmap is for that and how it works?
Dan Kearnan: HANA service pack 2 and 3, which will be released at the end of the year, will enable Business Warehouse 7.3 to run directly on top of HANA as a data layer.
The in-memory database of the data layer, where BW continues as the application layer, that helps you define your queries, your cubes, your hierarchies, et cetera. But the data layer which traditionally runs on Oracle or [IBM] DB2 could definitely be replaced by HANA as a new data layer, the in-memory database layer. That allows you to take the data that is traditionally stored in a database and have that data instead stored in the in-memory database running in conjunction with HANA.
There’s been some confusion over SAP HANA replacing SAP BW. Can you clarify that a bit?
Kearnan: There are no plans to ever take away or move from our strategy of BW as an enterprise data warehouse. Sometimes it will be a side-by-side scenario, where some companies will move forward with BW for enterprise data warehousing scenarios but will have HANA as a more nimble, flexible data environment for analytics that actually allow the kind of drill down into detailed operational transactional information with some degree of modeling, with a very quick turnaround time where business users get what they need.
Is there a lot of confusion around this question?
Kearnan: There is a lot of confusion. Before we had HANA, people used BW. They used it as a high-performance operational data mart, or some used it as a full enterprise data warehouse. They used it for everything. Now we’re giving them better options, for a high-performance analytic environment rather than trying to use BW for everything.
But HANA replaces SAP BWA, correct?
Kearnan: HANA will replace BWA going forward.
The next HANA service pack coming out at the end of the year will allow BW to run on top of HANA. Because BW will be running on top of HANA, customers that use BWA to speed up the performance of BW can instead look to HANA for that same function. They can look to HANA as being an accelerator for BW [to do the same thing that] BWA does.
So what are their options? Do they need to move to HANA?
Kearnan: They don’t have to move to HANA. They can run BW on traditional databases, and they can use BWA to speed up the performance of BW. We’re not taking that option away from them. They can continue to use and invest in BWA, and we will continue to enhance and maintain the BWA/BW strategy. Or, they can go to option two. They can move to HANA and adopt HANA as being the high-performance data foundation underneath BW and also, by default, have all the advantages of BWA that they’re used to and have come to know.
What kind of company makes the best kind of fit for HANA?
Kearnan: The industries are quite varied. Retail and CPG [consumer packaged goods] are good because of the data intensive nature of the industries: the financial sector, insurance companies. Other examples include the automotive industry.
One of the things [they share] when it comes to HANA is the large volumes of data they have in their ERP and transactional systems. Number two, the need to get real-time insight into this data because of the nature of the business, [and they need a more] granular approach to analysis.
What’s one thing that customers are confused about?
Kearnan: A lot of customers just think that if [they’re] not targeting an SAP ERP environment, then HANA isn’t for them. In fact, HANA has been designed from the ground up to be able to source any data, anywhere, whether you’re an SAP customer or not, as an independent, agnostic, high-performance data foundation. It’s an appliance that’s independent of the data source.