Creating a Custom Data Flow Component

In Microsoft SQL Server Integration Services, the data flow task exposes an object model that lets developers create custom data flow components—sources, transformations, and destinations—by using the Microsoft .NET Framework and managed code.

A data flow task consists of components that contain an IDTSComponentMetaData100 interface and a collection of IDTSPath100 objects that define the movement of data between components.

Note

When you create a custom provider, you need to update the ProviderDescriptors.xml file with the metadata column values.

Design Time and Run Time

Before execution, the data flow task is said to be in a design-time state, as it undergoes incremental changes. Changes may include the addition or removal of components, the addition or removal of the path objects that connect components, and changes to the metadata of the components. When metadata changes occur, the component can monitor and react to the changes. For example, a component can disallow certain changes or to make additional changes in response to a change. At design time, the designer interacts with a component through the design-time IDTSDesigntimeComponent100 interface.

At execution time, the data flow task examines the sequence of components, prepares an execution plan, and manages a pool of worker threads that execute the work plan. Although each worker thread performs some work that is internal to the data flow task, the principal task of the worker thread is to call the methods of the component through the run-time IDTSRuntimeComponent100 interface.

Creating a Component

To create a data flow component, you derive a class from the PipelineComponent base class, apply the DtsPipelineComponentAttribute class, and then override the appropriate methods of the base class. The PipelineComponent implements the IDTSDesigntimeComponent100 and IDTSRuntimeComponent100 interfaces, and exposes their methods for you to override in your component.

Depending on the objects used by your component, your project will require references to some or all of the following assemblies:

Feature Assembly to reference Namespace to import
Data flow Microsoft.SqlServer.PipelineHost Microsoft.SqlServer.Dts.Pipeline
Data flow wrapper Microsoft.SqlServer.DTSPipelineWrap Microsoft.SqlServer.Dts.Pipeline.Wrapper
Runtime Microsoft.SQLServer.ManagedDTS Microsoft.SqlServer.Dts.Runtime
Runtime wrapper Microsoft.SqlServer.DTSRuntimeWrap Microsoft.SqlServer.Dts.Runtime.Wrapper

The following code example shows a simple component that derives from the base class, and applies the DtsPipelineComponentAttribute. You need to add a reference to the Microsoft.SqlServer.DTSPipelineWrap assembly.

using System;  
using Microsoft.SqlServer.Dts.Pipeline;  
using Microsoft.SqlServer.Dts.Pipeline.Wrapper;  

namespace Microsoft.Samples.SqlServer.Dts  
{  
    [DtsPipelineComponent(DisplayName = "SampleComponent", ComponentType = ComponentType.Transform )]  
    public class BasicComponent: PipelineComponent  
    {  
        // TODO: Override the base class methods.  
    }  
}  
Imports Microsoft.SqlServer.Dts.Pipeline  
Imports Microsoft.SqlServer.Dts.Pipeline.Wrapper  

<DtsPipelineComponent(DisplayName:="SampleComponent", ComponentType:=ComponentType.Transform)> _  
Public Class BasicComponent  

    Inherits PipelineComponent  

    ' TODO: Override the base class methods.  

End Class  

See Also

Developing a User Interface for a Data Flow Component