After Microsoft’s announcement of SQL Server Data Services (SSDS) last week in Las Vegas, I have the feeling that some confusion has been generated. In fact sometimes when I speak with people, I often get the impression that they are thinking that SSDS is a Hosted SQL Server solution. But it’s really not that! Don’t think you can just upload your SQL Server database and start using it! Maybe the confusion is generated by the fact that in SSDS the name “SQL Server” is present.
Anyway as I said it’s not pure SQL Server in the cloud. In fact the type of data you can upload on SSDS depends on the below data model: you have entities scoped to containers and containers can contain different collection of entities. As you can see from the figure below taken from Nigel Ellis’s presentation at MIX (which I strongly suggest to see), Entities can be defined as property bag with a few fixed properties (ID, Kind, EntityId, EntityKind) grouped as metadata and a number of “open” properties (FlexProps) that can contain different types of data (string , Boolean, numeric, datetime). A container is the unit of consistency and scope of all operations.
Just another quick note: you can access SSDS by using internet standards based protocols (SOAP, REST), therefore you can access your data services from any platform.
If want to get more information go to the http://www.microsoft.com/sql/dataservices.