Ingest File Attachments from Salesforce CRM Objects into Data 360

How File Attachment Ingestion Works

When you set up file attachment ingestion, Data 360 ingests all files attached to CRM objects, regardless of when you created the connection to the CRM org and its data streams. The relevant data streams are:

• Content Document: Represents documents uploaded to a library in Salesforce CRM.
• Content Version: Represents the specific version of a document in Salesforce CRM.
• Content Document Link: Represents a link to users, groups, records, or libraries that the file is shared with.

You can ingest file attachments of the types listed under Unstructured Data File Formats

During setup, install a content bundle to deploy the three content data streams, and then create a search index to enable file attachment ingestion. The search index triggers an initial day-0 ingestion of up to 100,000 of the most recent historical files that were attached to the CRM object before you deployed the content bundle. This 100,000 limit applies to each CRM object, regardless of how many search indexes you create. After the initial ingestion, all new file attachments are incrementally ingested and relevant search indexes are automatically refreshed.

The ContentDocument, ContentVersion, and ContentDocumentLink data streams ingest metadata for all files. The 100,000 limit applies only to file content.

Note

If you've been ingesting file attachments from CRM to Data 360 prior to July 2026, the ContentVersion data stream behaves differently from other CRM content objects.

• ContentDocument and ContentDocumentLink (structured data) ingest all the content during the initial ingestion.
• ContentVersion (unstructured file content) ingests only five versions that you create or update after you create the data stream. To trigger historical file ingestion, delete the existing ContentVersion data stream and deploy the standard content_bundle. Then, either create new Salesforce objects with file attachments or use the existing attachments to the existing Salesforce objects.
• File attachment ingestion doesn’t support full refresh or total replacement for the ContentVersion data stream.

Before you set up a file attachment ingestion, plan your workflow by one of these scenarios.

Scenario	Steps
First-time setup	Use the Salesforce CRM connector to install the content_bundle that deploys the Content Document, Content Version, and Content Document Link data streams. Attach files to standard or custom Salesforce objects, and deploy a data stream for your desired CRM object. Create a Search Index with the appropriate source and attachment DMO.
You created content data streams before July 10, 2026	Any content data streams that you created before July 2026 aren’t compatible with this file attachment ingestion workflow. If you created the ContentDocument, ContentVersion, and ContentDocumentLink data streams before July 2026, you can only ingest files that were attached to CRM objects after you created the data streams. To ingest your historical file attachments, delete the existing content data streams and deploy the standard content_bundle. Then, either create new Salesforce objects with file attachments or use existing attachments to existing Salesforce objects.

Set Up File Attachment Ingestion

Step 1: Deploy a Data Stream for Your Desired CRM Object

Step 2: Deploy the Standard Content_Bundle

Note If you already have the Content Document, Content Version, and Content Document Link data streams, delete them first. Delete any downstream dependencies, such as search indexes or retrievers.

Install the standard Content_Bundle to deploy the Content Document, Content Version, and Content Document Link data streams. Learn more about installing standard data bundles powered by data kits.

1. From the app launcher, open Data 360 (Data Cloud).
2. In Data 360, go to the Data Streams tab.
3. Click New to create a new data stream.
4. Select data source: Salesforce CRM, and click Next.
5. Select the CRM org with the file attachments that you want to ingest, and click View Bundles.
6. Under Standard Bundles, select Content_Bundle and click Next.
7. On New Data Stream, make sure that for each data stream on the left, all the fields are selected. You don’t need to configure any other setting.
8. Click Next to continue.
9. Select the relevant data space and click Deploy. The content data bundle deploys these three data streams for your CRM object and their inter-relationships.
10. Go to your Data Streams tab to confirm the relationships are created and active.

    Content Document: Represents documents uploaded to a library in Salesforce CRM.

    Content Version: Represents the specific version of a document in Salesforce CRM.

    Content Document Link: Represents a link to users, groups, records, or libraries that the file is shared with.

Step 3: Create a Relationship Between the Desired CRM Object DMO and the Content Document DMO

After Data 360 deploys the three content data streams and their inter-relationships, configure the object relationship.

1. In Data 360, go to the Data Model tab.
2. From Data Model Objects, select the Content Document Relationships.
3. Open the Relationships tab and click Edit.
4. Add a new relationship row with these values.

   Object	Content Document Relationship
   Field	Related Item
   Cardinality	N:1
   Related Object	Data Cloud and the object that has file attachments
   Related Field	The primary key of the object that has file attachments, for example: Account Id

5. Save the new relationship.
6. Go to Data Model and then Data Model Objects and Content Document Relationship to ensure that a relationship exists between the object DMO (for example, Case, Opportunity, Account, or your custom object DMO) and the ContentDocumentRelationship DMO.
7. Confirm the following relationships also exist. Data 360 usually creates these relationships automatically during the Content_Bundle deployment. If one them doesn’t exist, create it manually using these values.

   Object	Content Document Relationship
   Field	Content Document
   Cardinality	ManyToOne
   Related Object	Content Document
   Related Field	Content Document Id

   Object	Content Document
   Field	Last Published Version
   Cardinality	ManyToOne
   Related Object	Content Version
   Related Field	Content Version Document Id

   Object	Content Document Version
   Field	ContentDocument
   Cardinality	ManyToOne
   Related Object	Content Document
   Related Field	Content Document Id

Step 4: Create Search Index with Advanced Setup

If you create the content data streams without an accompanying search index, Data 360 ingests only file metadata. To trigger file ingestion, create a search index. After you create it and select the correct attachment data model, Data 360 automatically detects and ingests historical file attachments in reverse chronological order (newest first).

Note For existing search indexes, Data 360 doesn't ingest the historical file attachments. To ingest historical file attachments, rebuild the search index.

1. From Data 360, open the Search Index tab and click new to create a new search index.
2. Click Advanced Setup, and then click Next.
3. Under Select Search Type, select Vector Search or Hybrid Search.
4. Select the relevant data space and the source data model object (DMO) where your file attachments reside.
5. On Select Fields to Chunk, click Include File Attachments.
6. Select the ContentDocumentVersion DMO, and click Save.
7. Configure other settings as needed and save the Search Index.

   After you create a search index, Data 360 immediately starts to ingest and index file attachments to make the contents available in the search index. To validate that Data 360 ingested and indexed your files, you can query the Chunk DMO via Query Editor and then confirm you see relevant chunks from your files.

   Data 360 refreshes the data as described in Data Refresh Process for Extraction. New files are included in this incremental ingestion. Data 360 captures any new file attachments or deleted files as events from the data streams.