At TechEd 2011 in Atlanta last month, Microsoft had a lot of weighty things to say about the cloud. During a keynote speech rife with announcements about the Windows Azure platform, the overarching message was this: The cloud's efficient; it’s economical; it’s the future.
So how does SQL Azure, Microsoft’s cloud database service, fit into so grand a plan? In this month’s “SQL in Five,” Microsoft database platform specialist Mark Kromer takes on that question, explaining some of the key SQL Azure developments unveiled at TechEd. He also gives a glimpse of what’s in the works for SQL Server, both on the ground and in the cloud.
How will the DAC Framework 1.1, announced at TechEd and due later this summer, make migration between on-premises SQL Server and SQL Azure easier? How does this update improve on existing import/export methods?
Mark Kromer: Today, you can build data-tier applications (DAC) in Visual Studio 2010 and deploy those applications to SQL Server 2008 R2 or to SQL Azure. This functionality is a new capability in SQL Server to provide database developers the ability to develop their schemas in Visual Studio (or in SQL Server Management Studio as well) and to create a self-contained “DACPAC” that can be used to separate the data from the schema and is very useful in providing version control, application lifecycle management and improving the communication and tracking of database schema changes between versions. It also makes for a much easier way to hand off developer changes to DBAs [database administrators] without needing to go through a backup and restore process to do so. The DAC feature is built on the DAC Framework (aka DAC Fx) and is being revised in the DAC Framework 2.0. Notice that this functionality is eliminating boundaries between SQL Azure and SQL Server so that you can use your DACPACs on SQL Server to migrate applications easily to SQL Azure. And the 2.0 version of the DAC Fx API [application programming interface] will include support for more data types (such as spatial data types) and better in-place upgrading.
Microsoft will ship DAC framework with SQL Server Denali. How does this fit into the company’s overall vision of SQL Server’s next release?
Kromer: There are a lot of changes taking place that are combining the features released for traditional SQL Server in conjunction with features in SQL Azure. This is an example of the synergies that Microsoft is forming by combining the development efforts of both products. So while the DAC Framework 2.0 looks like it will ship with all editions of Denali on Denali’s time schedule (late 2011/early 2012), SQL Azure has a CTP [community technology preview] available called the SQL Azure Labs import export tool, which demonstrates the DAC capability of both schema and data as well as the additional data types.
What is SQL Azure Management REST API and what advantage does it offer businesses?
Kromer: By leveraging the standard Web protocols in REST [representational state transfer], Microsoft has enabled developers and ISVs [independent software vendors] to leverage an API via REST to manage SQL Azure. Typically, managing a SQL Azure environment today is done either through the Windows Azure portal tools, which are based on Silverlight in the browser, or with the classic SQL Server on-premises tools like SSMS. But with the REST API, you can access your SQL Azure environment for automation or custom scripting of database management of SQL Azure databases.
What was the need behind Project Austin, which brings SQL Server’s StreamInsight complex-event-processing technology to the cloud? How will businesses benefit from it?
Kromer: The way that I see the Azure platform-based version of StreamInsight is similar to the other integration technologies in the cloud that Microsoft is releasing, such as AppFabric. Developers and data integrators today have a lot of options and architectural choices to make from the Microsoft stack when looking at integration. StreamInsight in its SQL Server 2008 R2 incarnation represents the best out-of-the-box capabilities around handling complex event processing such as what is needed to monitor networks for a NOC [network operations center] system when events may need to hold in a cache and only have meaning once a buffer detects that a threshold within a sliding window has been reached. In those cases with thousands and millions of message in a short period of time, you do not want to store them all in the database. StreamInsight provides that capability through Visual Studio and LINQ, and moving that to Azure is a natural progression when compared with the rest of the Microsoft development platform move into the cloud. I’m sure that SQL Server database and data warehouse developers will keep a close eye on this development to see what becomes of Microsoft Azure platform ETL [extract, transform and load] tools such as SSIS [SQL Server Integration Services] in terms of making tools like SSIS available in the cloud.
How do these SQL Azure developments work into Microsoft’s overall cloud strategy?
Kromer: The SQL Server and SQL Azure products are beginning to bring features to market in a manner that shows new synergies between the development teams. Some examples that you will see in the Denali release that overlap SQL Azure and SQL Server in terms of net-new features include contained databases, DAC v2 and multiple database replicas, aka Always On. In terms of the Microsoft Azure story, SQL Azure is key to the Microsoft strategy because it is the Microsoft public cloud database. SQL Azure is a market leader in terms of multi-tenant cloud databases. In fact, I have found that customers will often begin their journey of moving their IT platform into the cloud with SQL Azure because Microsoft offers the database as a separate part of the Azure platform that you can subscribe to.
Mark Kromer has over 16 years experience in IT and software engineering and is well-known in the business intelligence (BI), data warehouse and database communities. He is the Microsoft data platform technology specialist for the mid-Atlantic region. Visit his blog at MSSQLDUDE.