Crossing the Hadoop Chasm
While Hadoop as a driver of big data projects has been around for several years it’s just now that solution providers are starting to see the technology in mainstream productions environments. That’s because Hadoop itself is complex and many organizations have been deliberate in terms of finding actual use cases where Hadoop adds value to the business.
But that may be changing. Case in point — EMC has signed agreements with Hadoop-focused companies Cloudera and Hortonworks. Under the terms of these deals, EMC is now reselling both Cloudera Enterprise and the Hortonworks Data Platform while continuing to support a distribution of Hadoop created by Pivotal, an EMC sister company.
As Hadoop’s move into production environments accelerates, Tim Stevens, vice president of business and corporate development for Cloudera, told The VAR Guy that the provider of one of the two leading distributions of Hadoop is already seeing about half of its revenues driven by partnership alliances such as the one with EMC. Such agreements let companies such as Cloudera and Hortonworks more easily engage a small army of potential channel partners that it would take years for them to develop on their own.
And these deals give solution providers a way to quickly engage with companies like Cloudera and Hortonworks, through existing channel programs that EMC has in place to drive both Hadoop and related IT infrastructure.
Solution providers should still be careful about how they position Hadoop. There are basically two camps inside the Hadoop community. One sees Hadoop primarily being used as an extension to existing data warehouses that provide an inexpensive mechanism for storing massive amounts of data. The second sees Hadoop usurping existing data warehouses altogether. In that scenario, expensive data warehouses will eventually be replaced with new types of data warehouses built on a less expensive Hadoop platform.
Creating an Hadoop practice
To one degree or another both scenarios will come to pass in time. Right now most of the Hadoop adoption is being driven as an extension of existing data warehouses. But there’s also no doubt that given the chance most organizations would sorely love to sharply reduce the millions of dollars they currently spend on data warehouse applications and systems.
The most important thing for solution providers to remember at this juncture is that Hadoop is no longer a computer science project being conducted somewhere in the back of the data center. It’s becoming a strategic enabling technology that solution providers across the channel consider building a practice around as it continues to disrupt a data warehouse category worth billions of dollars.