carloscastilla - Fotolia

SAP HANA Hadoop enhancements called major step toward IoT

SAP HANA’s new Service Pack 10 release is geared to big data and remote synching across the enterprise with improved SAP HANA Hadoop integration.

SAP is moving to better integrate SAP HANA with big data and the Internet of Things (IoT) with the release of Service Pack 10 (SPS10), which includes enhanced support for SAP HANA Hadoop.

SPS10 for the SAP HANA platform delivers new capabilities that help enterprises connect to the IoT, manage big data more effectively, and extend the availability of data across the enterprise, according to the company.

One key aspect of the release is a new synchronization feature for data between the enterprise and remote locations across a network, said Marie Goodell, SAP's head of platform marketing. "Remote data sync allows two-way, session-based data replication between SAP HANA -- which can be located in the enterprise -- with remote databases that are out in distant places," Goodell explained. "Customers can now send data from the enterprise out to those remote locations, enabling customers and field workers to be more productive. Field workers can use their mobile devices to do things like inspections even if the Internet is offline because the data continues to move whenever the access is there. Also, the IoT data from the remote devices can come back to the enterprise where they can do additional analysis, which is important for areas like oil rigs, mining sites and any places where there's no high-band Internet."

Support for big data has been extended with the integration of HANA to the latest Hadoop distributions from Cloudera and Hortonworks. "Smart data integration allows you to access the data from Hadoop, load into HANA really rapidly and then do the analysis from there," Goodell said. "We've also added faster analysis data transfer with Spark SQL, and we've added a single unified user interface using the Apache Ambari so that administrators can look at HANA and the Hadoop clusters in one simplified administrative aspect."

SPS10 also delivers new high-availability and disaster-recovery capabilities that can ensure that mission-critical applications are always on and ready, Goodell said. Scalability has also been enhanced by allowing companies to use the non-uniform memory access (NUMA) architecture to support large-scale systems with more than 12TB of memory to process large data sets quickly. In addition, spatial processing now supports multidimensional objects and spatial expressions in HANA models or SQL Script.

SAP HANA Hadoop enhancements needed for IoT

Several aspects of the release show that SAP is building a solid foundation for HANA's growth, according to some analysts.

"It's interesting because HANA has been built very quickly and SAP was under tremendous time pressure, so this is a major housekeeping release doing lots of things that customers have been asking for," said Holger Mueller, vice president and principal analyst at Constellation Research. "The whole high-availability, disaster-recovery piece is actually pretty big. It may sound little on paper but in reality, in sales [there] can be major challenges to close a deal."

Mueller said many SAP users have been worried about losing data in an in-memory database if the power goes out. "SAP always had some OK answers -- but not great answers -- but now they are in significantly better shape," he said.

The enhanced SAP HANA Hadoop support, which had been tepid in the past, is essential for the message that SAP wants to send with HANA, Mueller said. "SAP was making an awful lot of talk about how they want to tackle IoT-based scenarios, but without Hadoop they could not play there," he said. "There's no enterprise software vendor in general -- it's not just limited to SAP -- that doesn't need to have a big-data, Hadoop-related strategy."

The release shows that SAP is serious about delivering an enterprise-class database environment, said Carl Olofson, a research vice president who covers data management software for IDC. "Things like the increased workload management and NUMA optimization for better scalability are very important," he said. "Multi-tenancy support is interesting because it shows they're working hard to make good that they're a cloud database. Along with the improved [high availability and disaster recovery], you know these are the things that are necessary developments."

SPS10 addresses some of the concerns that companies might have before they commit to HANA, which should make it a little easier to make such a huge change, said Olofson. "People who are part of the institutional framework resist change anyway, but especially when it's as radical as this," he said. "Running a database that's based in memory as opposed to one that's based on disk is a big change. It changes the hardware configuration, it changes the operations schedule, it changes the roles of people in the data center -- which is the big one -- and so you're going to get a lot of resistance."

Olofson said SAP's attempt to cover all the technical bases should make it easier for businesses to decide to go ahead with HANA.

Next Steps

Learn what was in HANA SPS09

Read a book excerpt on SAP HANA data storage

Understand other features of SAP HANA

Discover the role of HANA Vora in the SAP big data strategy

Dig Deeper on IoT APIs, Applications and Software