Work with datasets in n-tier applications


Datasets and related classes are legacy .NET Framework technologies from the early 2000s that enable applications to work with data in memory while the applications are disconnected from the database. They are especially useful for applications that enable users to modify data and persist the changes back to the database. Although datasets have proven to be a very successful technology, we recommend that new .NET applications use Entity Framework Core. Entity Framework provides a more natural way to work with tabular data as object models, and it has a simpler programming interface.

N-tier data applications are data-centric applications that are separated into multiple logical layers (or tiers). In other words, an n-tier data application is an application that is separated into multiple projects, with the data access tier, the business logic tier, and the presentation tier each in its own project. For more information, see N-Tier data applications overview.

Typed datasets have been enhanced so that the TableAdapters and dataset classes can be generated into discrete projects. This provides the ability to quickly separate application layers and generate n-tier data applications.

N-tier support in typed datasets enables iterative development of the application architecture to an n-tier design.It also removes the requirement to manually separate the code into more than one project. Start out designing the data layer by using the Dataset Designer. When you're ready to take the application architecture to an n-tiered design, set the DataSet Project property of a dataset to generate the dataset class into a separate project.