Microsoft Courts Big Data Market With SQL Server Update
Microsoft has released to manufacturing the newest version of its relational database, SQL Server 2012, the company announced Tuesday. The software will be available for customer purchase on April 1.
With this release, Microsoft is positioning SQL Server as a tool that can help organizations analyze large amounts of unstructured data, an increasingly popular practice called big data.
"With the continually decreasing costs of storage and compute, we've moved into an era where it is more affordable for organizations to store everything," said Doug Leland, Microsoft general manager of its business platform marketing group. "Of course, the challenge is how to derive insight from all these signals."
SQL Server 2012 is "a modern data platform that embraces the world of traditional structured data and brings in the world of unstructured big data," Leland said. The software is designed to help "customers keep pace with the change and help them evolve as their needs change."
The update will be the first major one to the database package since the 2010 release of SQL Server 2008 R2. The company has touted advancements in three broad areas for this release: better mission-critical readiness, better business intelligence tools and greater compatibility with cloud computing.
Additionally, with this release, Microsoft is promoting SQL Server as an essential tool for helping organizations execute big data analysis. SQL Server can act as a link between unstructured data platforms such as Hadoop and more traditional data warehouse-based business intelligence tools offered by Microsoft. Using SQL Server in conjunction with Hadoop, "customers will be able to derive insights from the data," using tools such as Excel PowerPivot and the SQL Server 2012's Power View, Leland said.
SQL Server 2012 has been modified to work with Microsoft's Azure-based deployment of Hadoop. Last October, the company launched, as a preview service, a Hadoop service on its Azure cloud. For the SQL Server 2012 release, availability of this Hadoop Azure service has been expanded to 2,000 users, from 400 in the initial launch. It plans a full release of the service by the end of June.
"Customers don't have to take on the costs of implementing Hadoop in the data center. They can sign up for the managed service offering," Leland said. The company also plans to issue a connector that would allow SQL Server to access Hadoop through Windows Server 8. It is also working on a Hadoop server distribution for Windows Server, which customers can use in their own data center, Leland said.
One early user of SQL Server 2012 for big data duties is social networking analysis service Klout. Klout uses Hadoop to process about 350GB of social networking data each day. The company uses Microsoft SQL Server Analysis Services (AS) to prep 600 million rows of data from Hadoop's Hive data warehouse component daily so they can be queried.
"We use AS as an analytics query engine since Hive and Hadoop are terrible at serving interactive queries at scale," said David Mariani, Klout vice president of engineering. "AS queries billions of rows in under 10 seconds and allows us to iterate and test a large number of scenarios very quickly."
Not everyone agrees that SQL Server will be an essential element of big data analysis. With this release, Microsoft is playing catch-up to the big data trend, said Curt Monash, of Monash Research. SQL Server will be a good fit for existing Microsoft shops, where the IT staff already understands how to work with the Microsoft technologies. It offers only a few unique features for non-Microsoft shops when compared with other relational offerings, however.
"Microsoft SQL Server is an adequate product if you don't mind being locked into the Microsoft stack," Monash said.
Microsoft will hold an online tutorial on March 8 to help potential users become more acquainted with the new features in SQL Server 2012.