BI Acceleration: Hortonworks HDP® 2.6 HIVE with LLAP - opening up new possibilities in EDW

By Alok Gupta

Remember hearing these statements?  

  • “Only a proprietary system can meet an enterprise’s very specific needs in such a complex area.”
  • “Nothing you can just go out and buy in the marketplace could ever deliver the same capabilities,”
  • “Sure, the price is a little … shall we say … um … high, and it doesn’t adapt easily to change, but it delivers what nothing else – and I do mean nothing else - can.”

 I do… starting years ago, applied to just about every area of IT. People would accept them as true for a while. Then, open systems solutions would show up and blow these claims out of the water.

That’s exactly what’s happening in the area of Enterprise Data Warehouse (EDW) today, thanks to the newly-introduced BI Acceleration: Hortonworks HDP® 2.6 HIVE with LLAP solution.

 Case closed?

Enterprises used to turn to proprietary DW systems for a very good reason: they made sure you got Fast BI and deep slice-and-dice analytics. But a number of not-so-good features came along with them, including unsustainably high initial prices, expensive maintenance contracts, and inability to evolve as data pattern and analytics requirements change. These solutions also have difficulty performing complex analytics and are inefficient with ETL/ELT tasks using 50% - 90% of CPU and interfere with more business critical tasks like BI and advanced analytics

 Cost-constrained enterprises, i.e., every single enterprise I know of, started asking the obvious questions. Would history repeat itself by creating an open source alternative? If so, when?

Thanks to HPE and Hortonworks, we have the answers. They are: “Yes” and “Now.”

 It’s like opening up a whole new EDW world

BI Acceleration: Hortonworks HDP® 2.6 HIVE with LLAP solution uses Hive LLAP to deliver Massive Parallel Process (MPP) performance at Hadoop scale. Enterprises get sub-second query performance for BI tools. At the same time, this new HPE Hadoop - Hortonworks solution is open-sourced and enterprise ready, making it easy for enterprises to unlock value from their existing Hadoop investments with a data centric infrastructure platform that meets their business needs. The solution also offers:

  • Two editions:  Standard Edition for customers who need balanced compute and storage AND Elastic Edition for customers who want to scale compute and storage independently.
  • Add-on Solutions to addresses security, data management and analytics needs
  • Complete services portfolio – everything from helping organizations build their Big Data transformation advisory roadmap to designing the right architecture to implementing the solution.
  • HPE Insight Cluster Management Utility enables customers to quickly provision, manage, and monitor their infrastructure.

It is important to note that the BI Acceleration: Hortonworks HDP® 2.6 HIVE with LLAP solution leverages key Hortonworks differentiators, including a centralized architecture built on YARN, resulting in: 

  • Shared services ensure consistent effective policies implemented for governance & security. Single point for operations streamlines deployment
  • Shared storage and processing creates efficiencies: less hardware, less cost
  • Consistent approach to on-boarding: new engines and new applications deployed inherit shared services.

You can add more resources to existing cluster without spinning up a new cluster

See Figure 1 below for a summary of features and benefits of BI Acceleration: Hortonworks HDP® 2.6 HIVE with LLAP solution. Figure 2 show the results of Hive LLAP vs Non-Hive LLAP Tests – HDP 2.6 for both the Standard and Elastic Editions of the solution. 


 Figure 1



Figure 2

Great closing arguments

Enterprise-grade Hadoop with LLAP solution does what opens source solutions have always done: blow proprietary approaches out of the water. You get an open source, cost effective and fast BI-centric data warehouse that can handle large volume of structured and unstructured data.

For complete information on Hortonworks Data Platform (HDP®) 2.6 running on HPE’s Enterprise-Grade Hadoop solution, contact your HPE Representative.  



0 Kudos
About the Author


HPE Alliance Partners