Friday, October 28, 2011

Oracle Embraces Hadoop and Big Data Within a Mainframe Construct

After dismissing Hadoop and NoSQL technologies earlier this year as mere topics of interesting academic discussion, Oracle used its high-testosterone Open World conference to announce its Big Data Appliance, a new addition to its family of engineered systems.

Give Oracle credit for cleverly embracing open source distributions of Hadoop and R without forsaking its engineered systems approach.

Make no mistake, the Big Data Appliance is no different than its "Exa" brethren -- a platform reminiscent of the IBM mainframe, optimized for a single, vertically integrated software stack.

It's more than ironic that Andy Mendelsohn, the Oracle executive who downplayed enterprise viability of NoSQL technologies just nine months ago at a financial analyst conference, would do the honors of unveiling Oracle's Big Data Appliance at this year's Open World. Particularly since it comprises a NoSQL database, and various ETL tools for Hadoop -- technologies, according to Mendelsohn earlier this year, that were entirely obviated by Oracle's Exadata Database Machine.

But it's clear Mendelsohn and his cohorts used the last nine months very productively to determine
how best to embrace Big Data open source technologies while simultaneously furthering Oracle's engineered systems strategy.

Mendelsohn proved the ideal choice to deliver Oracle's keynote on its Big Data strategy. He knows the subject well, and gave a cogent presentation on the life cycle of big data, and how an enterprise customer can harness various Oracle technologies to extract actionable intelligence to make better business decisions.

For those in the audience old enough to remember the mainframe, Mendelsohn made a comparison between NoSQL database key/value stores and VSAM on the mainframe.

The NoSQL-VSAM comparison develops further when one reads the news facts from Oracle's Big Data Appliance press release.

Simply put, Oracle's end-to-end solution to an enterprise Big Data problem requires purchasing three Oracle engineered systems: the Oracle Big Data Appliance in conjunction with the Oracle Exadata Database Machine and the Oracle Exalytics Business Intelligence Machine.

All three Oracle engineered systems are needed to acquire, organize, analyze, and maximize the value of Big Data within an enterprise.
This is Oracle's re-creation of the IBM mainframe in its own image.

It's a brazen strategy that clearly maximizes value and profits for Oracle. But it's unclear how much of the enterprise market is prepared to embrace Oracle's Big Data solution they claim works best within a mainframe construct.