BigQuery
Connect Google BigQuery for federated analysis. Datasets and tables are queried through BigQuery and made available to the Agent, and can be joined with datasets from other connectors.
When to use
- Analytics data warehoused in BigQuery.
- Cross-source analysis that combines BigQuery tables with an operational database or object storage.
Before you begin
Create a Google Cloud service account with read access to the target data:
- Roles: BigQuery Data Viewer on the datasets to be analyzed, and BigQuery Job User on the project so queries can run.
- Generate a JSON key for the service account. You will paste its contents into the connection.
Grant the narrowest dataset-level access that covers the tables you need.
Connection settings
| Field | Required | Description |
|---|---|---|
| Project ID | Yes | The Google Cloud project that runs the queries and is billed for them. |
| Location | Yes | The BigQuery location of the data, e.g. US or EU. Must match the datasets' location. |
| Service account JSON | Yes | The full JSON key of the service account. Stored encrypted; never shown after saving. |
Test and save
Enter the fields and use Test connection. On success, choose the tables to import. Datyo reads each table's structure and a sample of rows.
Capabilities
- Execution model: Federated. BigQuery tables can be joined with datasets from other connectors.
- Access: Read-only. Provide a viewer-level service account.
- Location-scoped: Queries run in the configured location; a mismatch with the data's location will fail.
Troubleshooting
- Authentication fails — confirm the JSON key is complete and valid, and that the service account has both Data Viewer and Job User roles.
- A dataset is not visible — grant the service account access to that specific dataset.
- Location mismatch — set Location to the region where the datasets actually reside.
- Unexpected query cost — BigQuery bills scanned bytes; scope analyses and prefer partitioned or clustered tables where available.