Monday, June 25, 2012

SQL Azure: Development Tools

There are number of tools available to build your SQL Azure database. In this post I will describe three of them including SSMS (SQL Server Management Studio), SSDT (SQL Server Data Tools), and the SQL Azure Database Manger.

1) SQL Server Management Studio
SQL Server Management Studio, a.k.a SSMS, and its predecessor Query Analyzer have been the tools of choice for SQL Server database developers that prefer a GUI over a command line. SQL Server 2008 R2 includes support for SQL Azure in SSMS. Naturally, the version of SSMS that ships with SQL Server 2012 also supports the management and development of SQL Azure databases.

While the SSMS 2012 user interface looks a bit different than previous versions because it is now hosted in the WPF-based Visual Studio shell, connecting to a database is the same as connecting from SSMS 2008 R2.

Once connected, the Object Explorer tree looks very similar to an on-premises SQL Server database, although you'll notice that the server icon looks a bit different. Additionally, some folders such as Server Objects and Replication as well as the SQL Server Agent node are missing as this functionality is not currently available on the SQL Azure platform. Refer to MSDN for a list of the SQL Server features not currently supported on the SQL Azure platform.   

Now that we have provisioned a SQL Azure server instance as well as created a SQL Azure database, let's create a table. Just like a regular SQL Server database, it's as simple as expanding the database node in Object Explorer, right-clicking the Tables folder and selecting New Table.... For some reason Microsoft has chosen not to provide the familiar table designer user interface. Instead it displays a CREATE TABLE script template. Below is a screen shot of a modified template used to create a PolicyTransactionFact table in our InsuranceDW database. All we had to do was name the table and add columns and data types. Once you have created your table definition you can either hit the F5 key on your keyboard or click the Execute button. The screen shot below was taken after executing the CREATE TABLE script.

Creating other types of SQL Azure objects is just as easy and is extremely similar to creating objects in on on-premises SQL Server database.

2) SQL Server Data Tools
SQL Server Data Tools (SSDT) provides a declarative approach to database development that is more in line with Microsoft's Visual Studio-based application development paradigm. It is hosted within the Microsoft Visual Studio shell and is free for download on MSDN.

 Microsoft would like SSMS to remain as an Administrative tool for SQL Server instances and databases but would like development work to be done in Visual Studio by way of SQL Server Data Tools projects. Additionally, starting with SQL Server 2012, SSDT replaces BIDS as the development environment for SSIS, SSAS, and SSRS. 

The first step in SSDT-based database development is to create a SQL Server Database Project. Make sure you install SSDT prior to attempting to create this project type as it is not available with the default installation of SQL Server 2010.

Once you've created the SQL Server Database Project right-click the top-level project node in Solution Explorer and select Properties. Here is were we set the Target Platform to SQL Azure to make sure that our build output is compatible with the SQL Azure platform.

Once the project is created and the platform appropriately set we can begin to create database objects. Just as we did with SSMS, we'll create a table. As you can see below, SQL Server Database Projects provide a table designer so that you can create tables using a GUI as well as writing SQL.

Since SSDT is a declarative development environment we don't actually create physical objects on a server as we develop them in code. Once the developer is content with their development they must Publish the current SSDT representation of the database schema to the database server. This is done by right-clicking on the top-level database project node and selecting Publish.

At this point we are presented with a dialog requesting the publishing/deployment information.

Once the connection information and database name are entered hit the Publish button. The results of the publish can be seen in the Data Tools Operations window.

SSDT SQL Server Database Projects are extremely powerful and can play an integral role in DLM (database lifecycle management) and ALM (application lifecycle management) processes. I provide an argument for using SQL Server Database Projects' predecessor Visual Studio Database Project in my blog post titled Visual Studio 2010 Database Projects and also explain the evolution to SQL Server Database Projects in my blog post titled The Future of Visual Studio Database Projects.

3) SQL Azure Database Manager
The last database development tool I'll discuss is the web-based SQL Azure Database Manager. To use the Database Manager you'll need to log into the Azure Platform Portal and navigate to the Database section by clicking on the Database Link.

Now navigate the menu tree and select the appropriate database server.
Now click the "Manage" link. 

This will take you to the Database Manager log on screen. Enter your credentials and log on. You should see  the screen below.

Click the "Design" link in the lower left-hand corner of the screen.

Notice that we're in the Tables section and that we see a list of our tables. Click "New Table" to create a new table.
The Create Table screen should be self-explanatory. You can add/remove columns, set their data types, select primary key columns, set nullability, etc. When you're done, click "Save" and your new table will be added to the database.

As you've seen, anyone can create a SQL Azure database (whether you want to pay or not) and use any one of the three tools we discussed to create and maintain database objects. SQL Azure isn't right for everyone, but for some organizations and situations it can be a very powerful tool. Hopefully this post will at least convince you to get your feet wet in case you find yourself in one of these situations so that you will have had enough exposure to SQL Azure to know what to do next. Good luck!

Wednesday, June 20, 2012

SQL Azure: Building your first Microsoft database in the cloud

We’ve all been hearing about the cloud for a while and if you’re exposed to Microsoft at all that means Azure. If you’re a database person that means SQL Azure. So what’s it all about? 

Without getting too wordy, it’s more or less a high-availability, hosted version of Microsoft’s SQL Server database. Want most of the benefits of SQL Server but don’t have the hardware or human resources to run it yourself? SQL Azure might be right for you. Maybe you just want to do a quick proof-of-concept or prototype but don’t want to take the time of standing up a SQL Server. SQL Azure might be right for you.

I think I’m probably like a lot of developers in that new technologies can be a little hard to grasp and sometimes even a little scary until I get my hands on it and see it in action. The purpose of this post is to show you how easy it is to create and use database in the cloud on Microsoft’s Windows Azure platform.

Many developers probably don’t know that they get a free subscription to the Azure platform with any Visual Studio Professional, Premium, or Ultimate with MSDN license. Additionally, BizSpark program members also get the same free subscription that is available with Visual Studio Ultimate with MSDN. Details can be found here. If you don’t fall into the MSDN or BizSpark camps and want to give Azure a test drive but don’t want to cut into your budget there is also a free 90-day trial available at the time of this post (

Time to start playing. At this point I’ll assume you’ve signed up, have a subscription and are logged in to the Windows Azure portal ( You should see the Getting Started with Windows Azure screen below.

To begin the process of creating a database server and database, click on the Database link on the bottom left hand corner of the screen.

Now click on the Create a new SQL Database Server link.

Choose your subscription. My is named POC for proof-of-concept. 

Select the Region in which you’d like your server hosted.

Now set a password.

You’ll now have to add a firewall  rule to allow incoming traffic from the machine(s) in which you’ll be managing this server.

Once you’ve completed the process of creating a server you’ll come to a screen that provides all the details of your server. You’ll also see that you already have one database, the master  database. The master database is required and cannot be deleted. This database contains things such as permissions and usage metrics.

Now that we have a server we can create a database. Click the Create link in the ribbon-like menu at the top of the screen. Provide a name for the new database and choose an appropriate edition and maximum size. To be safe (in terms of billing) choose a 1 GB Web Edition database.

Congrats! You’ve now successfully created your first Microsoft database in the cloud. 

Tuesday, June 19, 2012

SSIS Transactions Including Oracle Data Sources

Microsoft SQL Server Integration Services has the out-of-the box capability to handle transactions. I won't get into the details of how to use it as this post is specifically geared towards including an Oracle data source/Connection Manager in a transaction. For an overview of transactions in SSIS as well as enabling MSDTC to support this capability, please refer to the links below.

When using transactions with only SQL Server Connection Managers the typical gotcha for the uninitiated is the MSDTC requirement mentioned above. When using an Oracle-based Connection Manager there is an additional dependency. This dependency is the Oracle Services For Microsoft Transaction Server component and is available in the Oracle client installer. Below is a screen shot of the installation screen with this component highlighted. 

Once MSDTC and the Oracle Services For Microsoft Transaction Server are installed and enabled, cross-database platform transactions are a breeze.