Connected and Disconnected Architecture in ADO.Net

ADO.NET provides two primary data access architectures: connected and disconnected. These architectures define how an application interacts with a database and retrieves data.

Connected architecture

Connected architecture is a traditional way of accessing data. In this model, the application maintains a continuous connection to the database while it is reading or updating data. This means that the application must open a connection to the database, execute a query, and then close the connection. This can be inefficient, especially for applications that need to read or update large amounts of data. The key components of this architecture are the Connection, Command, and DataReader objects.

  1. Connection: The Connection object represents a connection to the database. It provides methods and properties to establish and manage the connection. The application needs to explicitly open and close the connection.
  2. Command: The Command object is responsible for executing SQL statements or stored procedures on the database. It allows the application to send queries and retrieve data from the database. It supports different types of commands such as SqlCommand for SQL Server or OracleCommand for Oracle databases.
  3. DataReader: The DataReader object provides a forward-only, read-only stream of data from the database. It allows the application to efficiently retrieve and process large amounts of data in a sequential manner. The DataReader is ideal for scenarios where the application needs to iterate over a result set and doesn't require the ability to modify the data.

Disconnected architecture

Disconnected architecture is a more efficient way of accessing data. In this model, the application reads data from the database into a disconnected data object, such as a DataSet or DataTable. The application can then work with the data in the disconnected object without having to maintain a connection to the database. When the application is finished working with the data, it can save the changes back to the database. The disconnected architecture offers several benefits:

  1. DataAdapter: The DataAdapter object is used to fetch data from the database and populate the DataSet or DataTable. It acts as a bridge between the application and the database. It supports operations like filling the dataset, updating changes back to the database, and managing concurrency.
  2. DataSet and DataTable: The DataSet is a container that holds multiple DataTables, which represent tables from the database. It can store multiple result sets and maintain relationships between tables. DataTables, on the other hand, represent a single table structure and can be modified independently.
  3. Data Manipulation: Once the data is retrieved and stored in the disconnected DataSet or DataTable, the application can manipulate, analyze, and update the data without being connected to the database. Changes made to the dataset can later be reconciled with the database using the DataAdapter.
When you might use disconnected architecture:
  1. An application that needs to read a large amount of data from a database and then process it offline.
  2. An application that needs to access data from multiple users or devices.
  3. An application that needs to improve performance by reducing the number of database connections.
When you might use connected architecture:
  1. An application that needs to read or update data in real time.
  2. An application that needs to access a database that does not support disconnected architecture.
  3. An application that needs to minimize the amount of data that is transferred between the application and the database.


ADO.NET provides connected and disconnected data access architectures to suit different application requirements. The connected architecture facilitates real-time access to the database, while the disconnected architecture offers the ability to work with data offline and perform extensive data manipulation before synchronization with the database. The choice of architecture depends on factors such as data access needs, performance considerations, and scalability requirements.