SQL Server 2012 SP1 Released
On Friday Microsoft released the first service pack for SQL Server 2012.
The news was delivered at the company's PASS event, held last week in Seattle. Corporate Vice President Ted Kummert introduced a number of demos showcasing existing and new technologies at work. He also gave an overview of Microsoft's overall direction in addressing business needs with its flagship relational database management system.
Key themes of the talk included access to self-service business intelligence (BI) through the use of user-friendly Excel-based tools, the use of "in-memory" technologies and the ability of organizations to tap unstructured data sources to gain business insights. Kummert said that Microsoft has "fully integrated Power View and PowerPivot into Excel 2013." According to Microsoft's self-service BI conception, knowledge workers can use those familiar tools to glean business insights from piles of data without bothering IT too much.
The keynote talk with the demos can be accessed on-demand here. The new service pack can be downloaded at this page.
SP1 for SQL Server 2012 adds some important capabilities that Microsoft had previously touted when it released the server back in April. For instance, the service pack enables "self-service BI as a natural part of users' day-to-day activities in Excel 2013." It adds the ability of users to mash up data from various sources using Excel's PowerPivot or create visualizations from data using Power View. It allows users to use Microsoft's xVelocity column store index to work with "hundreds of millions of rows of data." SQL Server Reporting Services improvements to support SharePoint 2013 also are available from this service pack.
SP1 supports cross-cluster migration of AlwaysOn availability groups. It adds a "selective XML index" capability, which is designed to improve querying XML data. Microsoft also added a way for IT pros to install SQL Server 2012 using a "prebuilt slipstream image," which simplifies the setup process. More SP1 improvements are listed in this TechNet library article.
Parallel Data Warehouse Arriving in 2013
Also announced at the keynote was the forthcoming availability of SQL Server 2012 Parallel Data Warehouse product, which Microsoft indicated would happen sometime in "the first half of 2013." This product will include a new "PolyBase" data processing engine, which is used to query Hadoop data, either from relational or nonrelational data sources.
Christian Kleinerman, a principal group program manager at Microsoft, showed a demo illustrating PolyBase. He said that if you know how to use T-SQL, then you just need to learn one new concept to tap the power of the PolyBase engine. IT pros need to create what's called "an external table." He ran a query that enjoins forms column table data with data from an order table.
"It's just another datasource for the query engine…. That's PolyBase," Kleinerman said.
He also showed speed improvements in Microsoft's current xVelocity column store index in-memory solution. In past demos, Microsoft clocked the counting of rows in wide tables at about 100 terabytes of data in two minutes. During today's demo, with a row count of 294 billion, SQL Server 2012 delivered 1 petabyte of data in one second. The speed improvement is because of Microsoft's xVelocity technology, according to Kleinerman. The xVelocity technology also provides storage savings. It can shrink the space needed by up to ten times, Kleinerman claimed.
Future "Hekaton" In-Memory Technology
Microsoft's "next major version of SQL Server" will include a new in-memory transactional technology called "Hekaton." If an organization has applications that need to run faster, then Hekaton can provide a solution, said Shawn Bice, director of program management of the Database Systems Group at Microsoft. Hekaton converts tables so they run in memory. Hekaton will be built into the next SQL Server and it won't require hardware or software upgrades to run, according to Microsoft. The Hekaton solution is currently being testing by online gambling company bwin. It's only available in a "private technology review with a small set of customers," according to a blog post by Kummert.
Also demonstrated during the keynote was the use of Microsoft's BI tools to gain insight from unstructured data. Amir Netz, a Technical Fellow on the Microsoft BI team, showed how to gain insights about movie star branding using unstructured Twitter data. He used Microsoft's new HDInsight connector solution to collect 12 million Twitter tweets about movie stars and charted it according to the publicity associated with the Academy Awards. Some stars, like Brad Pitt, did not have big spikes of interest by the time of the Academy Awards because he's generally popular, according to Netz's analysis of the data. He used Power View, part of SharePoint Services, to chart the information over time.
Microsoft previously announced previews of HDInsight Server for Windows and the Windows Azure HDInsight Service last month. HDInsight uses a bidirectional connector, which allows data to be moved between Hadoop and SQL Server 2012 or SQL Server 2012 Parallel Data Warehouse in order to analyze structured data.
Kurt Mackie is senior news producer for the 1105 Enterprise Computing Group.