Skip to content
Create a data source › Data sources › Import sets › Imports › Workflow Data Fabric
Release: Australia · Updated: 2026-03-12 · Official documentation · View source

Create a Data Stream (Integration Hub) type data source

Create a Data Stream (IntegrationHub) type data source record to define what complex data an import set should import from a Data Stream action.

Before you begin

Role required: import_admin and action_designer

Procedure

  1. Navigate to All > System Import Sets > Administration > Data Sources.

  2. Click New.

  3. Complete the form.

FieldDescription
NameA unique name for this data source.
Import set table labelA label for the import set staging table to use.
Import set table nameThe ServiceNow AI Platform uses the label you entered to construct a unique table name. This prevents namespace collision with an already existing table.
TypeData Stream \(Integration Hub\) type data source.
Data Stream action
  • Image omitted: select-data-stream-action.png
    Select Data Stream action Selected to use an existing published Data Stream action from Integration Hub.</li>
  • Image omitted: create-data-stream-action.png
    Select published Data Stream action or create new Data Stream action Selected for one of the following reasons:</li>
  • If you select a published Data Stream action, click this button to open the Data Stream action in Integration Hub.
  • If you don't select a Data Stream action, click this button to create and publish a new Data Stream action in Integration Hub. Note: Data Stream actions require a subscription for production use.
ConnectionSpecify the connection alias for the Data Stream action. You can use the default parent alias or select any of its children's aliases. Available child aliases are listed in the menu. This field is only available when the Data Stream action is configured to use a connection alias. For more information, see Create a Connection and Credential alias.
Data in single columnSelected to store data in a single column.
Use Batch ImportIf selected, specifies that loading data should be batched when you select Load data or Test load 20 records.Note: The system keeps records in memory until it receives the number of records specified in the batch size. If the batch size is met, it commits all the records as one batched SQL. If all the records are loaded for the import set, it commits all the remaining records as one batched SQL. Since it keeps the maximum of the batch size records in memory when specified, make sure you set the correct batch size to not cause any out-of-memory issues.
Batch Size

Specifies how big the batches should be. Appears only when the Use Batch Import field is selected.The glide.import_set_load_usebatch and glide.import_set_load_batch_size properties affects these fields. The priority order for the properties and fields are as follows: If the property is set, use the property value. If not, use the field value. If neither property is set, use the default batch size of 1000.

  • For example, if glide.import_set_load_usebatch = true and glide.import_set_load_batch_size = 200, use 200 as the batch size, regardless of the field values.
  • For example, if glide.import_set_load_usebatch = true and the glide.import_set_load_batch_size property is not set, use the default batch size of 1000.
  • For example, if the glide.import_set_load_usebatch property is not set and the Use Batch Import field = true, use the Batch Size field value (or the default of 1000 if the Batch Size field value is not set)
  • For example, if the glide.import_set_load_usebatch property = false and the Use Batch Import field = true, do not use batch import, since the property is set to false.
Data Stream action inputsAdditional fields may appear, based on the inputs defined in the Data Stream action.
  1. Click Submit.

What to do next

Define a transform map for the new data source and run an import.