Project Kraken is a joint initiative by Hewlett-Packard and SAP to help enterprises improve their business processes through super high-speed reporting and data analytics. A prototype for the new system, which uses HP hardware and SAP
technology, was demonstrated Thursday at the Sapphire Now conference in Orlando, Fla. The initiative is based on HP server technology innovations and optimized with the in-memory SAP HANA database for the SAP Business Suite.
HANA sounds like an acronym, but is actually a name coined by SAP AG for its database technology. In basic terms, the benefit of HANA database technology is the unusual speed and power it provides by keeping all data in-memory at all times.
Using that method, processes like data analysis and reporting can work much faster than if the system had to move data between the database and outside applications. The time savings can be critical for big-data environments that require massive processing power and immediate system response times.
More Data, Less Time
With Project Kraken, HP and SAP say they are on a mission to give customers more choice in scalability for large, online transaction-processing applications that require real-time analytical data insights. If the technology performs as marketed, it will give retailers, financial institutions, utilities, governments, and others a new alternative to detect patterns, analyze massive data volumes on the fly, and perform their operations quickly.
"With Project Kraken, HP and SAP are illustrating how customers can achieve a step-jump in performance, while adding simplicity in the management of the environment," said Vishal Sikka, a member of SAP's Technology and Innovation executive board. "It shows how large enterprises can confidently run their applications, including SAP Business Suite, with growing quantities of data in shorter windows of time."
Sikka said the offering will "fundamentally and forever change the database market."
So what's under the hood? The prototype system is optimized with 16 Intel Xeon processor E7 CPUs (codenamed Ivy Bridge-EX) and 12 terabytes of memory.
The system has been engineered specifically for complex, data-intensive workloads, such as supply chain, customer relationship management (), enterprise resource planning (ERP) and data analysis. The technology makes it possible to consolidate multiple online transaction processing (OLTP) and online analytic processing (OLAP) databases into one vertically scalable server with improved performance.
SAP said Project Kraken will further the capabilities of HANA, providing real-time analytics on transactional data, eliminating the need to spend time and resources moving data between environments. HP and SAP are working to create a marketable solution that can scale to the highest levels of memory within a single server for speeding business operations.
Charles King, principal analyst at Pund-IT, spoke with us from the SAP conference. He explained there are nine server-vendors building systems based on SAP HANA -- and they all have roughly the same capabilities. Five vendors are SAP-certified to build scale-out systems. IBM is one, and he said it looks like HP is another.
"SAP announced last week a HANA enterprise that they say can scale out to 100 nodes and a full petabyte of data. So there's a linear scaling performance that HANA has been theoretically capable of," King said. "Kraken is one solution that sounds like its getting ready to come to market."
From what King has seen, the idea of blending traditional transaction processing and analytics processing is something that HANA is fully capable of supporting. That makes the partnership a positive for HP.
"Overall, I've been very impressed with the way that SAP has been leveraging its partnerships around HANA," King said. "The company has developed a pretty large footprint in the analytics space without having to buy a hardware company in order to do it. Kraken is a good example of the forward progress that both companies have generated."