February 14, 2013

Microsoft Pushes Hadoop up Windows Hill

Nicole Hemsoth

Back in November we took a look at how Microsoft was tapping into the open source momentum behind Hadoop with their take on the framework. At the time, the best analogy would be a small team pushing a giant elephant up a steep incline.

Even at the beginning of this year the project was still somewhat disorganized in terms of approach, in part because the team at Microsoft that had been working on a Hadoop competitor, Dryad, dropped the project in favor of working with the native Apache platform. This took some tying in of external expertise from Hortonworks, and quite a shift away from the company’s usual built-in-house approach.

More recently, they’ve continued their collaboration with companies like Hortonworks and SiSense to push their big data vision forward. At the core of Microsoft’s vision for big data is HDInsight, which is their own supported Apache-compatible distro for Windows Server and inside their Azure cloud.

While it’s still a little early to get a sense of real customer adoption at this point, Microsoft is proclaiming its commitment to the open source community (and their own Hadoop expertise) by beefing up its collaboration with Hortonworks, which has been steady since last year—and working with so-called big data analytics companies like SiSense to demonstrate their efficiency for enterprise users.

According to Herain Oberoi, who leads product management efforts at Microsoft, the two companies are still working together to move Hadoop further into Windows while pushing the fruits of their labor out into the open source community for better future compatibility.

According to Bruno Aziza from SiSense, this ability to use a big data platform that plays nicely with Windows infrastructure tools is one step toward “democratizing big data” and allowing users to process data quickly and let them scale horizontally.

 One of their users that has come forward is reputation ranking company, Klout, which found the SQL Server angle an attractive element in their decision to find a workable BI solution. More specifically, the company needed to expand to meet the terabyte-scale needs of users, thus they settled on a BI solution that was based on SQL Server 2012 Enterprise and Hadoop. This let them process queries in near real-time and according to Klout, saved them some dough in the process.

While we expect Microsoft to add a number of big name companies to its ranks of HDInsight users, it seems that they are still off to a slower start than others who managed to grab the Hadoop horse by the mane as it just galloped from the open source stables. The Windows element, however, will be attractive for many firms seeking the familiar SQL environment, although other companies have beat them to that game in the process as well.

Related Articles

New Ropes for Scaling the SQL Wall

SiSense Expands Big Data Cloud Offering

How Data Analytics is Shifting Politics