Power BI now defaults to ADBC driver for Databricks connections
The Apache Arrow-based ADBC driver is now generally available and the default for new Power BI connections to Databricks, streamlining columnar data transfers.
O que mudou
The Databricks connector for Power BI now uses the Arrow Database Connectivity (ADBC) driver by default for new connections. Reaching general availability (GA), ADBC provides a modern, columnar API standard based on Apache Arrow to transfer analytical data much more efficiently.
Use cases
Imagine your Power BI reports need to pull substantial volumes of data from a SQL Warehouse. The ADBC driver optimizes this transfer by bypassing the row-based translation bottlenecks inherent to the legacy ODBC driver.
You can use it to:
- Ingest analytical tables natively using columnar Apache Arrow format.
- Upgrade existing dashboards from ODBC to the modern ADBC implementation.
- Connect directly to Lakehouse//RT warehouses (Beta) via the REST-based Statement Execution API.
How to use
New connections created through the user interface use ADBC automatically. To switch existing reports that were built with ODBC, update the query in the Power Query Advanced Editor.
let
Source = DatabricksMultiCloud.Catalogs(
"meu_workspace.cloud.databricks.net",
"/sql/1.0/warehouses/meu_sql_warehouse_id",
[Catalog=null, Database=null, EnableAutomaticProxyDiscovery=null, Implementation="2.0"]
),
powerbi_Database = Source{[Name="meu_catalogo",Kind="Database"]}[Data],
default_Schema = powerbi_Database{[Name="meu_schema",Kind="Schema"]}[Data],
dashboard_data_Table = default_Schema{[Name="minha_tabela",Kind="Table"]}[Data]
in
dashboard_data_Table
What happens in this code:
DatabricksMultiCloud.Catalogstargets your workspace hostname and SQL Warehouse HTTP path.- Setting
Implementation="2.0"tells Power BI to execute requests via the ADBC driver instead of the legacy driver. - The navigation steps drill down into the target catalog, schema, and table to load data.
If you need to connect to a Lakehouse//RT warehouse, include Protocol="rest" in the options record.
Limitations and requirements
- Minimum versions: requires Power BI Desktop 2.145.1105.0 or higher. Connecting to Lakehouse//RT requires version 2.156.951.0 or higher.
- Single driver constraint: you cannot mix drivers inside the same semantic model. Every Databricks connection within a single model must be set to either ADBC or ODBC.
- Existing connections: pre-existing reports remain on ODBC until manually updated to
Implementation="2.0". - Power BI Service with Lakehouse//RT: direct connections in the Power BI service do not work because the service tests connections over ODBC; you must route through an on-premises data gateway and select Skip test connection.
- power bi
- adbc
- odbc
- apache arrow
- bi connector
ADBC driver for Power BI is now generally available
Resumo e explicação gerados por IA a partir da release note oficial. Em caso de dúvida, a documentação oficial da plataforma prevalece.
Relacionadas
- DatabricksBeta
Conector do SurveyMonkey chega ao Lakeflow Connect no Databricks
O Lakeflow Connect agora conta com conector nativo para o SurveyMonkey em Beta, permitindo ingestão incremental de pesquisas, respostas e contatos com governança no Unity Catalog.
lakeflow connect · surveymonkey - DatabricksBeta
Lakeflow Connect adiciona conector do Oracle Eloqua em Beta
O Lakeflow Connect agora oferece um conector gerenciado em Beta para o Oracle Eloqua, permitindo ingestão incremental de dados de marketing governados pelo Unity Catalog.
lakeflow connect · oracle eloqua - DatabricksBeta
Ingira métricas do GA4 com o novo conector do Lakeflow Connect
O Lakeflow Connect ganha conector gerenciado em Beta para o Google Analytics 4, permitindo ingestão incremental de relatórios agregados via API com governança no Unity Catalog.
lakeflow connect · google analytics