It appears Hadoop has yet another fan: Microsoft.
ZDNet is reporting that Microsoft is putting LINQ to HPC, its distributed runtime and associated programming model for unstructured data analysis at scale, on ice. Codenamed Dryad, LINQ to HPC was Microsoft’s proprietary Big Data framework meant to compete with open source Hadoop.
The move comes a month after Microsoft announced a deal with Hortonworks to deliver its Hadoop distribution, Hortonworks Data Platform, on the Windows Azure cloud platform. In her report, ZDNet’s Mary-Jo Foley cites a recent blog post from the Microsoft Big Data team explaining its change of direction. It appears, from my reading, that Microsoft simply decided it couldn’t keep up with the pace of development we’re seeing from the open source Hadoop community to compete. And I’m sure Microsoft crunched the numbers and figured there’s more money to be made playing off Hadoop than competing with it head on.
Microsoft is just the latest mega-vendor to validate Hadoop as the real-deal, got-to Big Data approach. Oracle announced at OpenWorld last month it would distribute its own Hadoop distribution running on its new Big Data appliance. IBM, meanwhile, is leveraging Hadoop as the foundation of its Big Insights platform fro large-scale unstructured data analytics. And HP has forged a hardware partnership with Cloudera, the leading commercial Hadoop vendor, while its Vertica division was the first MPP data warehouse to build a native Hadoop connector.
When Microsoft, Oracle, HP and IBM line-up behind you, you know you’re on to something. In both Microsoft’s and Oracle’s case, the two vendors plan to deliver Hadoop and related services from the cloud, exposing Hadoop to a potentially entirely new constituency.