Hewlett-Packard is betting $50 million that its customers will find value in the Hadoop data processing platform, by investing in Hadoop distributor Hortonworks.
The two companies will also partner on sales calls and formulate an architecture for incorporating the Hortonworks’ Hadoop distribution into HP’s data processing stack, called HAVEn.
“We’ve never really sold Hadoop, nor supported it directly. To date, we’d meet with a Hadoop vendor and take the deal to market. What’s different about this strategic partnership is that we’re now committing to resell the Hortonworks Data Platform as part of HAVEn, and support it as well,” said Colin Mahony, HP general manager for the company’s Vertica analytic database management software.
When implementing a large-scale data processing platform, enterprises would prefer to have a single vendor to deal with, rather than multiple vendors that each contribute a piece to the overall system, Mahony said.
In addition to the investment and joint engineering work, HP Chief Technology Officer Martin Fink will join Hortonworks board of directors , so the two companies can work closely when formulating Hadoop strategies.
First developed at Yahoo, Hadoop provides a way to store massive amounts of unstructured data that can be analyzed on the fly. It has found a home in many large Internet services, which collect so much data from users that it would be infeasible to store and analyze that data through the traditional technologies of SQL databases and commercial data warehouses.
One of the chief distributors of Hadoop, Hortonworks was founded by a number of engineers who worked on the original Yahoo implementation of Hadoop. Since its creation in 2011, Hortonworks has formed partnerships with Microsoft, SAP and Red Hat, among other enterprise software vendors.
The partnership will give Hortonworks a broader potential customer base, said Hortonworks CEO Rob Bearden.
The partnership will fill out the HP’s HAVEn stack. HAVEn stands for Hadoop, Autonomy, Vertica and enterprise security. The “n” stands for any number of enterprise applications that can be built on top of the data processing platform.
HP will also start working on connectors to better pair Vertica with the Hadoop YARN (Yet Another Resource Negotiator) management console. HP already has software that allows users to run Vertica commands against data stored on Hadoop systems and this work will ensure that Vertica’s view of Hadoop data will stay fully synchronized with what the Hadoop console sees.