Skip to main content
Feedback

Sources

note

Knowledge Hub is an Early Access release. Refer to the Early Access disclaimer for details.

A Source represents a single data connection between an external system and a Knowledge Base. It holds the configuration that tells Knowledge Hub where data comes from, how it is structured, and when to ingest it.

How a Source is created

A Source record is created when you deploy a source-to-target (S2T) flow in Data Integration with Knowledge Hub as the target. You can start that flow from either side:

  • From Data Integration: Build a new S2T flow, select Knowledge Hub as the target, and select the Repository and Knowledge Base by name.
  • From Knowledge Hub: On the Sources tab of the Knowledge Base, select a Data Provider (for example, Boomi Data Integration), then select Create Source. This opens Data Integration in a new tab to build the S2T flow, with the Repository and Knowledge Base already selected. Select Refresh on the Sources tab to see the deployed Source appear.

When the flow deploys:

  1. Data Integration sends a configuration snapshot to Knowledge Hub.
  2. Knowledge Hub creates a Source record in Draft state.
  3. The Source appears on the Sources tab of the Knowledge Base.

You then enable the Source from Knowledge Hub to start ingestion. Refer to Configuring Knowledge Hub as a target and Managing source lifecycle.

Sources tab

The Sources tab lists every data source connected to the current Knowledge Base. For each Source, you can see:

ColumnDescription
TitleThe Source name, generated automatically from the connector type and flow name when the Data Integration flow deploys.
Data ProviderThe provider that supplies data to this Source. Select the link to open the corresponding flow in Boomi Data Integration.
Source TypeHow data is delivered to Knowledge Hub (for example, Manual Upload).
Last IngestionThe timestamp of the most recent ingestion execution and its status (success or error icon).
Error SummaryA brief description of the error if the last ingestion run failed. Shows no content when the last run succeeded.
StateThe current state of the Source: Draft, Enabled, or Disabled.

From the Sources tab, you can enable a Source, trigger a manual run, disable a Source, or remove it. Refer to Managing source lifecycle.

What you need to complete the ingestion flow

Before you can see active Sources on the Sources tab, you must complete steps in both Knowledge Hub and Data Integration.

Step 1: In Knowledge Hub

Create a repository and a Knowledge Base. These are the ingestion target that Data Integration writes to. Refer to Creating a repository and Creating a Knowledge Base.

Step 2: In Data Integration

Set up a source-to-target (S2T) flow with Knowledge Hub as the target. Configure the source connector, schema, data preparation, and schedule, then deploy. You can start this flow directly in Data Integration, or select Create Source from the Knowledge Base's Sources tab to open Data Integration with the Repository and Knowledge Base already selected. Refer to Configuring Knowledge Hub as a target and the Data Integration documentation.

Step 3: Back in Knowledge Hub

After the flow deploys, a Source record appears on the Sources tab in Draft state. Enable it to start ingestion.

note

If no Sources appear after deploying in Data Integration, check the Data Integration execution log for deploy errors. A failed sync means Knowledge Hub did not receive the flow configuration.

Multiple Sources in one Knowledge Base

A Knowledge Base can have multiple Sources. All Sources write into the same vector index. There is no limit on the number of Sources per Knowledge Base in Early Access.

Common patterns:

  • Multi-connector ingestion: One Source from S3 (unstructured documents) and one from Confluence (structured wiki content) feeding one Knowledge Base.
  • Multi-environment ingestion: Separate Sources for a production database and an archive, contributing to a unified search index.

Source states

A Source moves through three states: Draft, Enabled, and Disabled. Archiving is not currently supported; to permanently remove a Source, use Remove. Refer to Asset states reference for the full state model, or Managing source lifecycle for state-change procedures.

Disabled is not yet available for Knowledge Bases in Early Access. Once available, disabling the parent Knowledge Base will not change a Source's state, but ingestion runs for that Source will fail while the Knowledge Base is disabled.

Next steps

On this Page