Skip to main content
BI Tools integrations are enabled per organization for migration projects. Ask your Datafold contact or support@datafold.com to turn them on.
A BI Tools integration reads the definitions of your dashboards, workbooks, and semantic models, and proposes them as a pull request to your migration repository. With the definitions in the repository, a migration project can plan and track the move of your dashboards to the new warehouse. To add one, go to Settings → Integrations → BI Tools and click Add new integration. Then choose your BI tool. Only organization admins can add BI integrations.

Tableau

Power BI

​
What Datafold reads

Datafold reads definitions only: workbook and data source XML from Tableau, and semantic model (TMDL) and report (PBIR) definitions from Power BI. It does not download extracts, cached data, or query results. It also records an inventory of each site or workspace: item owners, links, update times and, for Tableau, view counts.

​
Settings shared by both tools

  • Name: the name of the integration. When you create the integration, Datafold derives its own folder, bi/<tool>/<name>/, from the name. A later rename does not move that folder, and each integration writes only its own folder, so several integrations can share one repository.
  • Git integration: the repository that receives the pull requests. It must be a GitHub (Datafold app), GitLab, or Azure DevOps integration, or AWS CodeCommit.
  • Base branch: the branch that the pull request targets. Empty means the repository’s default branch.
  • Folder in repository: a prefix for the bi/ directory, so that the sync writes to <folder>/bi/<tool>/<name>/. Empty means the repository root. The pull request can also update a sync summary under .datafold/ at the repository root.
  • Auto-merge pull request: squash-merge each sync’s pull request once its checks pass. Failing checks stop the merge, and branch protection still applies, so a required review keeps it open. Not available for CodeCommit.
  • Mappings: the Datafold data connection that each warehouse connection in your BI tool reads from. Run a sync first: Datafold then lists the warehouse connections that it found, and you pick a data connection for each. You can also add a mapping by hand. Datafold writes the mappings to connections.yml in the integration’s own folder.

​
Sync

Click Sync now on the integration’s page to run a sync. Datafold reads every site or workspace in scope and opens or updates one pull request on a branch that it owns. Do not push to that branch: Datafold rebuilds it on every sync. The Latest sync panel shows the status, the number of definitions and extract failures, and a link to the pull request (or No changes to propose). You can also follow syncs on the Jobs page. When one item fails to download, the sync keeps the repository’s previous copy of that item instead of treating it as deleted. A site or workspace that Datafold cannot read at all keeps what the repository already holds. A site or workspace that you remove from the integration’s settings is emptied in the repository on the next sync.

​
Network access

Datafold must reach your BI tool. If your Tableau Server or network restricts access by IP address, allow Datafold’s egress addresses (see Securing connections).