Business Software is in the business of data management. Creating, updating, deleting and using information in new ways as well as simplifying how it is collected and reported is key to what we do as technology professionals. With the advent of Cloud Computing being able to take advantage of relational data directly in the cloud using the same skills we’ve been using with solutions built on premise means we’ll spend less time on the good old learning curve. One great example where this is the case is SQL Azure, the relational database service from Microsoft.
If you’ve been working with SQL Server for any length of time you’ll know that at its base it is a relational database engine that talks on port 1433 and speaks a language called Tabular Data Stream (or TDS). Common tools for working with SQL Server on premise includes SQL Server Management Studio (SSMS), BCP, Integration Services, Visual Studio, and many more too numerous to name. What is interesting about SQL Azure is that it uses the same protocols and ports as SQL Server you run on premise, so all the same tools continue to work. In this post I’d like to show how to get started working with SQL Azure quickly, provisioning a server, creating a database and some schema, and then migrating data to it.
First thing we need is a service running in a datacenter to host our databases from. This is a logical service that represents many physical machines. We provision a server by going to the Windows Azure management portal (http://windows.azure.com) and selecting the Database section. From there we select our subscription and then click the Create button in the ribbon.
This will provision a SQL Azure service in a specified data center. The management tool lets you select the data center and set administrative credentials. These credentials are whatever you want, but with a couple restrictions. For example it disallows certain usernames for the admin account including SA, Root, Admin, Guest, etc. Next it requires a strong password which includes 3 of the 4 types of characters on the keyboard (Upper Case, Lower Case, Numbers, and/or Symbols). The last step is to set up some firewall rules which specify who can see the server. By default there are no rules included, you even need to check the box to allow Windows Azure Services to see the server. You should also add a rule for your development machine, but be aware that it’s looking at the external facing IP address. Fortunately when you click the Add button it includes your external facing IP address.
Once this is done you’ll see you have a new database service created. The server name is auto generated for you and is part of the *.database.windows.net domain. It includes a master database (which is used by the system, no charge to you), and you can add databases as needed. To create a database select your server and then click the “Create” button.
You’ll see a dialog that prompts you to decide the size of database and the name. You have options for Web and Business and sizes ranging from 1 GB to 150 GB.
After creating the database the ribbon provides buttons which include functionality such as management of the schema and performance data from your database, as well as tools to import and export the data and schema. You can also drop the database and/or the server from the ribbon.
In the right pane you’ll find properties about your database. The pieces of information that are most useful include the allocated size, actual size, management URL (which is simply the generated name + .database.windows.net), and a link button to see examples of the connection strings you’ll use in your application to connect to SQL Azure.
So that’s it. I’ll go into the web based management tool in another post, but the key is that it is very easy to get up and started working with SQL Azure. In less than 5 minutes you can provision a service and be up and running with your database in the cloud.