Updating Data Sources with DataAdapters
Important This document may not represent best practices for current development, links to downloads and other resources may no longer be valid. Current recommended version can be found here. ArchiveDisclaimer

Updating Data Sources with DataAdapters (ADO.NET)

The Update method of the DataAdapter is called to resolve changes from a DataSet back to the data source. The Update method, like the Fill method, takes as arguments an instance of a DataSet, and an optional DataTable object or DataTable name. The DataSet instance is the DataSet that contains the changes that have been made, and the DataTable identifies the table from which to retrieve the changes. If no DataTable is specified, the first DataTable in the DataSet is used.

When you call the Update method, the DataAdapter analyzes the changes that have been made and executes the appropriate command (INSERT, UPDATE, or DELETE). When the DataAdapter encounters a change to a DataRow, it uses the InsertCommand, UpdateCommand, or DeleteCommand to process the change. This allows you to maximize the performance of your ADO.NET application by specifying command syntax at design time and, where possible, through the use of stored procedures. You must explicitly set the commands before calling Update. If Update is called and the appropriate command does not exist for a particular update (for example, no DeleteCommand for deleted rows), an exception is thrown.


If you are using SQL Server stored procedures to edit or delete data using a DataAdapter, make sure that you do not use SET NOCOUNT ON in the stored procedure definition. This causes the rows affected count returned to be zero, which the DataAdapter interprets as a concurrency conflict. In this event, a DBConcurrencyException will be thrown.

Command parameters can be used to specify input and output values for an SQL statement or stored procedure for each modified row in a DataSet. For more information, see DataAdapter Parameters (ADO.NET).


It is important to understand the difference between deleting a row in a DataTable and removing the row. When you call the Remove or RemoveAt method, the row is removed immediately. Any corresponding rows in the back end data source will not be affected if you then pass the DataTable or DataSet to a DataAdapter and call Update. When you use the Delete method, the row remains in the DataTable and is marked for deletion. If you then pass the DataTable or DataSet to a DataAdapter and call Update, the corresponding row in the back end data source is deleted.

If your DataTable maps to or is generated from a single database table, you can take advantage of the DbCommandBuilder object to automatically generate the DeleteCommand, InsertCommand, and UpdateCommand objects for the DataAdapter. For more information, see Generating Commands with CommandBuilders (ADO.NET).

You can control how the values returned from the data source are mapped back to the DataTable following a call to the Update method of a DataAdapter, by using the UpdatedRowSource property of a DbCommand object. By setting the UpdatedRowSource property to one of the UpdateRowSource enumeration values, you can control whether output parameters returned by the DataAdapter commands are ignored or applied to the changed row in the DataSet. You can also specify whether the first returned row (if it exists) is applied to the changed row in the DataTable.

The following table describes the different values of the UpdateRowSource enumeration and how they affect the behavior of a command used with a DataAdapter.

UpdatedRowSource Enumeration



Both the output parameters and the first row of a returned result set may be mapped to the changed row in the DataSet.


Only the data in the first row of a returned result set may be mapped to the changed row in the DataSet.


Any output parameters or rows of a returned result set are ignored.


Only output parameters may be mapped to the changed row in the DataSet.

The Update method resolves your changes back to the data source; however other clients may have modified data at the data source since the last time you filled the DataSet. To refresh your DataSet with current data, use the DataAdapter and Fill method. New rows will be added to the table, and updated information will be incorporated into existing rows. The Fill method determines whether a new row will be added or an existing row will be updated by examining the primary key values of the rows in the DataSet and the rows returned by the SelectCommand. If the Fill method encounters a primary key value for a row in the DataSet that matches a primary key value from a row in the results returned by the SelectCommand, it updates the existing row with the information from the row returned by the SelectCommand and sets the RowState of the existing row to Unchanged. If a row returned by the SelectCommand has a primary key value that does not match any of the primary key values of the rows in the DataSet, the Fill method adds a new row with a RowState of Unchanged.


If the SelectCommand returns the results of an OUTER JOIN, the DataAdapter will not set a PrimaryKey value for the resulting DataTable. You must define the PrimaryKey yourself to ensure that duplicate rows are resolved correctly. For more information, see Defining Primary Keys (ADO.NET).

To handle exceptions that may occur when calling the Updatemethod, you can use the RowUpdated event to respond to row update errors as they occur (see Handling DataAdapter Events (ADO.NET)), or you can set DataAdapter.ContinueUpdateOnError to true before calling Update, and respond to the error information stored in the RowError property of a particular row when the update is complete (see Handling DataRow and DataColumn Errors (ADO.NET)).

Note   Calling AcceptChanges on the DataSet, DataTable, or DataRow will cause all Original values for a DataRow to be overwritten with the Current values for the DataRow. If the field values that identify the row as unique have been modified, after calling AcceptChanges the Original values will no longer match the values in the data source. AcceptChanges is called automatically for each row during a call to the Update method of a DataAdapter. You can preserve the original values during a call to the Update method by first setting the AcceptChangesDuringUpdate property of the DataAdapter to false, or by creating an event handler for the RowUpdated event and setting the Status to SkipCurrentRow. For more information, see Merging DataSet Contents (ADO.NET) and Handling DataAdapter Events (ADO.NET).

The following examples demonstrate how to perform updates to modified rows by explicitly setting the UpdateCommand of a DataAdapter and calling its Update method. Notice that the parameter specified in the WHERE clause of the UPDATE statement is set to use the Original value of the SourceColumn. This is important, because the Current value may have been modified and may not match the value in the data source. The Original value is the value that was used to populate the DataTable from the data source.

Private Sub AdapterUpdate(ByVal connectionString As String)

    Using connection As SqlConnection = New SqlConnection( _

        Dim adapter As SqlDataAdapter = New SqlDataAdapter( _
          "SELECT CategoryID, CategoryName FROM dbo.Categories", _

        adapter.UpdateCommand = New SqlCommand( _
          "UPDATE Categories SET CategoryName = @CategoryName " & _
           "WHERE CategoryID = @CategoryID", connection)

        adapter.UpdateCommand.Parameters.Add( _
           "@CategoryName", SqlDbType.NVarChar, 15, "CategoryName")

        Dim parameter As SqlParameter = _
           adapter.UpdateCommand.Parameters.Add( _
           "@CategoryID", SqlDbType.Int)
        parameter.SourceColumn = "CategoryID"
        parameter.SourceVersion = DataRowVersion.Original

        Dim categoryTable As New DataTable

        Dim categoryRow As DataRow = categoryTable.Rows(0)
        categoryRow("CategoryName") = "New Beverages"


        Console.WriteLine("Rows after update.")
        Dim row As DataRow
        For Each row In categoryTable.Rows
            Console.WriteLine("{0}: {1}", row(0), row(1))
    End Using 
End Sub

If the tables from your data source have auto-incrementing columns, you can fill the columns in your DataSet either by returning the auto-increment value as an output parameter of a stored procedure and mapping that to a column in a table, by returning the auto-increment value in the first row of a result set returned by a stored procedure or SQL statement, or by using the RowUpdated event of the DataAdapter to execute an additional SELECT statement. For more information and an example, see Retrieving Identity or Autonumber Values (ADO.NET).

In many circumstances, the order in which changes made through the DataSet are sent to the data source is important. For example, if a primary key value for an existing row is updated, and a new row has been added with the new primary key value as a foreign key, it is important to process the update before the insert.

You can use the Select method of the DataTable to return a DataRow array that only references rows with a particular RowState. You can then pass the returned DataRow array to the Update method of the DataAdapter to process the modified rows. By specifying a subset of rows to be updated, you can control the order in which inserts, updates, and deletes are processed.

For example, the following code ensures that the deleted rows of the table are processed first, then the updated rows, and then the inserted rows.

Dim table As DataTable = dataSet.Tables("Customers")

' First process deletes.
dataSet.Update(table.Select(Nothing, Nothing, _

' Next process updates.
adapter.Update(table.Select(Nothing, Nothing, _

' Finally, process inserts.
dataAdpater.Update(table.Select(Nothing, Nothing, _
© 2016 Microsoft