5 min read
_Editor's Note: Join SAP HANA Vora and MapR in a complimentary webinar January 19, 2017._
MapR and SAP share a common belief and vision about making big data analytics enterprise- ready, both at the platform and compute engine layer. In that regard, we are pleased to integrate and support SAP HANA Vora 1.3 and believe that it offers a complete solution for all types of big data analytic use cases. As you probably know, Vora is based on Spark and needs an underlying Spark platform to support these different analytic use cases. MapR has been a leader in offering Spark as part of the MapR Data Platform since the early days of Spark; in fact, MapR has shipped 8 releases of the full Spark stack since 2014.
The 1.3 release of SAP HANA Vora represents a new direction for big data analytics with the vision of having a single in-memory compute solution that delivers insights across a variety of data types including JSON, time-series, graph, and relational. The benefit to this approach is that multiple data types can be accommodated within the same data modeler, which implies a simplified architecture. We are also pleased to see SAP HANA Vora 1.3 ship with disaster recovery capabilities at the compute engine layer. Metadata persistence is enabled by a distributed transaction log that can be recovered when needed. This allows for recovery across the entire data model, in case of failures. For disaster recovery, the MapR Platform provides capabilities such as mirroring, replication, and consistent, point-in-time snapshots, which allow for fidelity and protection of critical data. SAP HANA Vora 1.3 also ships with new capabilities around enterprise-grade security. The MapR Platform supports unified security through capabilities such as granular access controls, and other key data management features such as volumes and snapshots.
Vora 1.3 also expands on its integration with SAP HANA. Hadoop integration has been a capability for a while in SAP HANA; Vora is bi-directional—so what does that exactly mean?
In summary, this means that you can expand the scope of analytic workloads and unify SAP and non-SAP data with Vora.
And if you weren’t aware, MapR can also provide primary storage for SAP HANA at a lower cost compared to alternatives, and enable greater flexibility. As a customer, this provides you with significant benefits—storage for HANA, and analytics from both SAP HANA and Hadoop, which can both be run on a single cluster. This is a unique differentiator for MapR.
There’s nothing better than illustrating a use case and showing how the two products are coming together to solve real-world problems. A large hi-tech company is enhancing their sales and marketing efficiency using Vora on MapR. In this scenario, their revenue and transaction data is stored in SAP HANA, with clickstream data being stored in the MapR Data Platform. They are looking to use Vora to analyze data from both these sources, improve the effectiveness of their marketing campaigns, and achieve better targeting to upsell and cross-sell customers. This is helping drive top-line revenue for them. In the future, they are also looking at the potential use of Vora on MapR in their supply chain function to be able to do more fine-grained bill of material analyses across several thousand products and product families. Other use cases we see for Vora on MapR include trading analysis to ensure optimal execution of trades in the banking space. Another use case is predictive maintenance to detect anomalies in real time in manufacturing and oil & gas, especially with the growth in IoT- enabled devices.
A sample architecture for SAP HANA Vora and MapR is shown below, which includes Vora analyzing data from both SAP HANA and the MapR Platform.
In our next post, we’ll walk through a tutorial to install and configure SAP HANA Vora 1.3 on MapR. You’ll then be able to get to a final configuration where you have Vora services running on a MapR cluster, as shown below.