The cloud first SQL Server 2014 coming in April with in-memory and cloud capabilities

microsoft-sql-server-2012

Two years after the arrival of SQL Server 2012, which focused on the analysis of data in memory and the cloud, Microsoft is ready for next iteration of the RDMBS database. The general availability of SQL Server 2014 has been set for April 1st by Microsoft, the eve of the opening of its Build developers conference which runs from April 2 to 4 in San Francisco.
While SQL Server 2012 was decidedly focused on high availability, data analysis and cloud storage, SQL Server 2014 focuses on performance and hybrid cloud. Microsoft has announced that the latest version of its SQL Server database has now been released to manufacturing (RTM), offering in-memory transaction processing to speed performance and features to support hybrid cloud deployment.

In-memory transaction processing

While the vast majority of its competitors have opted for in-memory technologies and opting it to separate from traditional RDBMS, Microsoft chooses to integrate OLTP technologies at the heart of its processing. The ambitious initiative that took four years of hard development has significant advantages.

Microsoft says SQL Server 2014 is much faster than its predecessor–at least in the online transaction processing (OLTP). The main reason is that OLTP is now equipped with an in-memory technology, whereby a speed improvement can be achieved by a factor of 30. For this purpose, Microsoft has developed an engine called Hekaton, which is developed by the SQL Server team in collaboration with Microsoft Research Database Group.

Microsoft indeed shares the principle that it is neither useful nor appropriate to seek to put absolutely all the databases in memory. In addition, by integrating this functionality at the heart of the engine, all enterprise applications can benefit from the resulting acceleration without having to rewrite or modify tables.

Microsoft uses as example an online gaming site called Bwin, which can process up to 250,000 requests per second with a pilot application. The new in-memory-based OLTP complement existing in-memory technologies for business intelligence (BI), Complex Event Processing with StreamInsight, and the column-oriented processing in SQL Server.

Another tool that comes standard with Microsoft’s offering can automatically analyze database activity and identify tables that would benefit resident in memory function. The built-in SQL Server Management Studio tool AMR (Analysis, migrate and report) will help the user determine if the in-memory OLTP can effectively improve the performance of your application. It analyzes the workload and indicates which tables and stored procedures that can take advantage of a conversion memory and the amount of work it will require. A Memory Optimization Advisor wizard then helps turn tables, some incompatibilities to be removed to be able to move in memory.

“This means our customers don’t need expensive additional software to take advantage of the technology. More importantly, it means customers do not have to rewrite their application or deploy new servers,” said Quentin Clark, Corporate Vice President, Data Platform Group. “We approached in-memory columnstore the same way – it is built into SQL Server and our data warehouse product Parallel Data Warehouse.”

However, the In-Memory OLTP functions are present only in the Enterprise edition and will not be available in other editions.

photo credit: EJP Photo via photopin cc

photo credit: EJP Photo via photopin cc

Hybrid cloud

The other major new focus has been on hybrid features that span the cloud and on premise compute, making it simpler for customers to back up and recover SQL Server databases hosted in their data center using Microsoft’s Windows Azure public cloud. This allows the new SQL Server optionally use for on premise Azure applications.

Windows Azure and SQL Server provide a continuum of capability and flexibility in deployment options both on-premises and in the cloud. Windows Azure also supports an important data workload – Windows Azure HDInsight, the Apache Hadoop-based solution in the cloud.

Apart from this, the backup mechanism to Windows Azure has been simplified. Backups can be done either automatically or manually in the future and be restored within minute in a Windows Azure Virtual Machine. Microsoft has announced that SQL Server 2014 Virtual Machines on Windows Azure would be available from 1 April.

SQL Server 2014 also provides many other improvements, including improved data management index column mechanism (an evolution of the current index store data in columns of SQL Server 2012), the ability to use SSD as buffer space for I/O, save databases on premise to Azure Storage via SSMS console integration, as well as many improvements in safety.

Microsoft is defining a strategy to create and update applications taking account of market trends namely cloud computing, mobile and big data. Microsoft most recent products release are developed in conjunction with Azure cloud platform. Last year, the software major released a cloud-based version of Visual Studio that offers developers to launch several new services running on the Windows Azure cloud platform.