SEATTLE -- The way Microsoft sees it, “big data” isn’t too big to take on.
It’s the way a lot of other vendors see it too. This week Microsoft joined the ever-lengthening queue of vendors -- HP, EMC Greenplum and IBM among them -- that hope to harness the waves of fast-moving, enormous sets of structured and unstructured data that are overwhelming enterprises by linking the upcoming SQL Server release and the Windows Azure cloud platform to big data workhorse Hadoop.
Microsoft announced the news Wednesday at the PASS Summit 2011, the annual Professional Association for SQL Server conference held here this week.
“The is the new world of data,” said Ted Kummert, senior vice president of the business platform division, describing the rise in data volume and data types as businesses rake in newer and more diverse types of information, such as image files and online data feeds. A data platform designed to tame that world and make sense of it, he said, must embrace traditional data centers and public and private clouds.
Apache Hadoop is the open source distributed computing framework that can process and manage data sets that are too large for traditional tools. To shape that relationship, Microsoft has partnered with Hortonworks, a Yahoo spinoff that provides training and support for Hadoop users to ultimately “grow the Hadoop ecosystem,” according to Eric Baldeschwieler, Hortonworks’ CEO. SQL Server’s multitude of users will help move Hortonworks closer to its goal, he said.
Microsoft’s vision is just as expansive. “We look at SQL Server as the foundation of delivering and managing any data, any size, anywhere,” said Doug Leland, general product manager for SQL Server.
The company’s first foray into Hadoop came this summer, with a tech preview offering connectors for SQL Server and Parallel Data Warehouse, allowing integration with existing environments. Final versions of those connectors are now available.
And with new capabilities comes a new name for the next version of SQL Server, known since its unveiling last year by its towering code name Denali. SQL Server 2012 will be released to manufacturing in the first half of next year, Kummert said. And with it will come PowerView, the official name of the visualization and reporting tool introduced as Project Crescent. Kummert said touchscreen capabilities, designed for use with tablets or smartphones, will be available by the end of next year.
“It’s no longer just a desktop or laptop world,” he said.
A tech preview of Hadoop-based Windows Azure service will out before New Year’s, and another for Windows Server will follow next year.
Though the words “big data” might be new, Leland said Microsoft has been in the business of large data sets for ages, noting, for example, that the company was the first vendor of online analytical processing technology, with SQL Server Analysis Services.
Last week Oracle, often seen as more aligned with the larger companies that would typically have a need for managing data on such a scale, announced its own Hadoop entry at its annual OpenWorld conference, further tightening the competition.
For Mark Tabladillo, who holds Microsoft’s Most Valued Professional distinction, Redmond is at a good starting point.
“They are well-positioned,” said Tabladillo, a database architect and founder of data mining and business intelligence consulting firm MarkTab Consulting. “To have the scalability of SQL Server is the evidence base for their position.”
Plus, Hadoop can help all businesses looking to gain new insights, Tabladillo said. “Having Hadoop as an option will give people the opportunity to get information that doesn’t fit in rows and columns.”
But there is a caveat, he said. “Smaller companies do have to make an exclusive choice, and Microsoft is hoping that choice will be SQL Server.”
But Richard Stockhoff, of consultancy RS Tech in Issaquah, Wash., saw Hadoop as Microsoft’s way of biting off a larger chunk of the market. “It’s expanding to bigger business,” he said, something that’s likely to influence everybody else. “Small and medium-sized businesses will eventually go into Hadoop and that will expand their market share.”
Also impressing Stockhoff was PowerView, which was designed to make it easy for end users to get reports exactly how they want them. And that’s a boon for availability, he said. “End users like to mess things up.”