dbt Extension

Keep the following in mind as you set up the dbt extension:

• The extension supports dbt Core v1.7.
• You can use Snowflake, Databricks, Redshift, Postgres, and BigQuery as Reverse ETL sources.
• dbt models aren't synchronized from the dbt cloud. The model sync connects to a Git repository that loads models into Segment for use with Reverse ETL.
• You can connect to GitHub using a GitHub App, token, or SSH.
• For GitLab and Bitbucket, use SSH to connect.

To set up the dbt extension, you'll need:

• an existing dbt account with a Git repository
• for job syncs, dbt cloud with jobs already created
• a user with Workspace Owner permissions in Segment

To connect Segment to your dbt models, you'll first need to configure a Git connection. This allows Segment to pull dbt models directly from your repository.

1. In your Segment workspace, navigate to Settings > Extensions.
2. Click either Set up dbt model syncs or Set up integration
3. On the Configure service credentials page, select a credential and protocol, add your key or token, then click Next.
4. In the Connect source window, select an existing Reverse ETL warehouse source from the dropdown, then click Save.

After you've saved your setup, you can configure your settings to your needs by changing the repository, branch, dbt version, default schema, and project path.

You can also use dbt Cloud to schedule Reverse ETL syncs after a dbt Cloud job successfully runs.

To set up dbt Cloud:

1. In your Segment workspace, navigate to Settings > Extensions.
2. Click Manage dbt Cloud.
3. Add your dbt Cloud API key or dbt Personal Access Token and an optional custom subdomain, then click Save.

The dbt Cloud integration allows you to schedule Reverse ETL syncs based on a dbt Cloud job. When a dbt Cloud job is selected under the Reverse ETL scheduling section, Segment creates a webhook in the dbt Cloud account that will initiate to run the Reverse ETL sync when the job is scheduled.

In order to create the webhook, ensure that you have webhook permissions associated with the dbt Cloud token in the previous step.

After you set up dbt, Segment runs an initial sync to load models from your connected Git repository. This initial sync lets you use the most recent models when you set up Reverse ETL. In addition to Segment's initial dbt sync, you can also trigger manual dbt model syncs.

After you've successfully set up dbt with a warehouse and connected to your Git repository, you can select dbt models for use with Reverse ETL by following these steps:

1. In your Segment workspace, navigate to Connections > Sources and select the Reverse ETL tab.
2. Click +Add Reverse ETL source , select your source, then click Add Model.
3. Click dbt Models as your modeling method, then select and preview a model from the dbt model dropdown.
4. Add a primary key, then click Preview your model.
5. Click Next.
6. Enter your Model Name, then click Create Model.

To change a connected model, ensure that you've removed it from all active Reverse ETL syncs.

Git Connections enable Segment to sync data with your preferred Git repository through supported like SSH and token-based authentication.

Segment supports the following credential types for setting up a Git Connection:

• SSH: Compatible with GitHub, GitLab, and Bitbucket, SSH provides a secure method for connecting to your repository.
• Git token: Git tokens are supported across GitHub, GitLab, and Bitbucket, enabling token-based authentication for added flexibility.
• GitHub App: For GitHub users, GitHub App integrations offer enhanced security and functionality. This method is exclusive to GitHub and supports additional features, like CI checks.

Segment lets you set up multiple Git Connections, allowing you to reuse credentials across both dbt and Git Sync. You can either use the same credential for multiple configurations or create separate Git Connections for each product and environment as needed.

If you plan to reuse a Git token across both dbt and Git Sync, ensure it has the necessary read and write permissions for both integrations.

CI checks in Segment help prevent breaking changes to active dbt models. Avoid changing dbt models currently in use with an active Reverse ETL sync, since changes could disrupt existing mappings and active syncs.

When CI checks are enabled, Segment monitors model changes in your Git repository. If a model already linked to an active Reverse ETL sync gets modified, Segment automatically rejects the change to maintain data integrity.

To enable CI Checks, authorize a GitHub App credential for your Git connection. Once connected, you can enable CI Checks in the dbt model sync configuration section.

The following table lists common dbt Extension errors, as well as their solutions: